Home  Resources  References  Tubes  Forums  Links  Support 
Zenith H511 no power
2/9/2008 1:12:00 PMJoe G
Hello out there,
I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
Thanks, Joe G.
2/9/2008 1:25:46 PMJon
:Hello out there,
: I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
:Thanks, Joe G.

Joe,
Check for voltage at switch input and out. You could have a faulty switch contact.
Jon.
:

2/9/2008 4:14:15 PMJoe G
::Hello out there,
:: I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
::Thanks, Joe G.
:
:Joe,
:Check for voltage at switch input and out. You could have a faulty switch contact.
:Jon.
::Jon, Switch has power in and out. Also picked up output for volume. Thanks,Joe
2/11/2008 10:05:03 PMEdd



Possibly, I can throw in a bit more.

With the initial assumption that the tube filaments and the dial light are lighting up. Also, that the unit was working in a fashion before you changed any parts. That certainly leaves a miswire or a connection error most suspect.

Since you mentioned schematicum unfamiliaritis , I am attaching a mark-up working schema, where you can just track down to its specified test points.

Schema insert:





Initially take a power down and do an ohmming out of the main B+ line [ marked up in red squares] which originates at the 35W4 pin 7 [Reference box A]and travels down to the left lead of surge inrush resistor R11...so ohms wise, you should only be reading wire resistance and test lead wire resistance, a very low ohmic value.

Move the probe over to the right of R11 [Reference box B] and then be expecting in the order of ~22 ohms, and then move down to the 60 ufd electrolytic and confirm the same resistance at its pos terminal, then probe up to the right side of R10, above, for the same resistance.

The sets main B+ buss then routes up to the power input side of the audio output which will probably not be terminal accessible, so move on past it and read for continuity at the plate ( left un-numbered by me, since I was unsure which AF out tube was being used.... pin 3 if it is a 50L6.

That should read a cumulative ~180 + 22 ohms total.

If all checks out, go back down to the R10 1000ohm voltage dropping / filter resistor which drops that main ~110Vdc down to a sub B+ level of ~ 93 V [ marked up in blue squares] and is started at [ Reference box C].

It makes 4 branch offs to feed circuits, the first one would be a direct run up to the screen grid of the AF output tube.

Take the negative ohmmeter lead and keep it on [ Reference box C] and use the other probe to check out [ Reference box C], where another low reading would be expected for a direct wire linking.

Then you move the probe over to pin 5 of the 12BE6 [ Reference box F] and then to pin 5 of the 12BA6[ Reference box G] , where you should expect ~ 20 ohms resistance from IF transformers primary windings being read out in the series path of either tube.

The last leg to be measured is going to encounter a 470K plate resistor in line, so when metering the plate of the 12AT6, expect ~470k resistance at [ Reference box H] . That should be enough initial continuity of B+ checks.

Move the ohmmeters neg ground lead to the black test box marked Gnd and then take the probe and check for continuity over to the two black G box references that are extensions of the initial gnd buss...low ohmmages..right ? as that should only be wire interconnects....if nothing was disrupted in parts change outs.
I really would suspect the trouble to be in the area of the filter capacitors wiring or the associated R10 and R11 resistors interconnecting wiring, where you might have made a miss wire or left something unconnected.

As far as what all parts you consider being involved in a "recapping" of that set, I would be initially considering the filter condensers and then look at the green mark up boxes for the marked capacitors. The number one most critical , being the heavy marked up one at the 1st grid of the AF output tube ( unless Zenith used a disc ceramic type for it...as well as disc ceramics being used in other places also.)[That type is a very,very, very trouble free capacitor.]

I also don't know if your unit used the little U1 couplate instead of the individual discrete res and caps between the 12AT6 and the output tube circuitry.

If C3 is using a paper type of cap, it might be a 'BERRY good idea to replace it with a modern day MKT series of capacitor that is safety rated for use directly across an AC line. When a conventional paper cap is used there, all it takes is for a high spike voltage from the AC line / and-or / a lightning hits induced voltage coming in to present itself across the caps electrodes and a punch through of the dielectric . That initial plasma path then permits the whole 120 volt line voltage continually dump its fury across that conductive path and then there is fireworks until that path burns open enough or the cap explodes.

If it seems that the power supply is operative with voltages, you then might try a testing of the power supply and the AF output stages by taking a test lead with clips at its ends and take a spare .001 thru -.1 ufd capacitor and place one test lead end clip on pin 4 of the 12AT6 to extract a 6 Vac test signal and take the other free end of the test lead and clip on to one end of that capacitor, with the other cap lead used to initially / sequentially touch to the 1st grid of the AF output tube…expecting hum if all is well…. and then move the free cap lead up to the 1st grid of the 12AT6 expecting much more hum and then finally turn the volume control to max and and move the cap over to inject signal into terminal 4&5 at the bottom of the secondary of the 2nd IF transformer. All of these are sequentially numbered on the markup as [1] [2] [3].

Standing by for feed back..... as the next step might be powering up and checking for the tubes screen and plate voltages.




73's de Edd





:::Hello out there,...can you hear me now ?
::: I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
:::Thanks, Joe G.
::
::Joe,
::Check for voltage at switch input and out. You could have a faulty switch contact.
::Jon.
:::Jon, Switch has power in and out. Also picked up output for volume. Thanks,Joe

2/12/2008 8:29:16 PMDoug Criner
Agree with Edd. Everything points to a wiring error. It's easy to do - right in the middle of replacing a component, the door bell rings, the wife calls, the phone rings, etc. Then, when you come back, it's easy to forget things.

It's a little late to suggest this, but I always try a set after each component is replaced. If the set isn't working intially, I troubleshoot first.
Doug

:
:
:
:
:
:Possibly, I can throw in a bit more.
:
:With the initial assumption that the tube filaments and the dial light are lighting up. Also, that the unit was working in a fashion before you changed any parts. That certainly leaves a miswire or a connection error most suspect.
:
:Since you mentioned schematicum unfamiliaritis , I am attaching a mark-up working schema, where you can just track down to its specified test points.
:
:
:
:Schema insert:
:
:
:
:
:
:
::Initially take a power down and do an ohmming out of the main B+ line [ marked up in red squares] which originates at the 35W4 pin 7 [Reference box A]and travels down to the left lead of surge inrush resistor R11...so ohms wise, you should only be reading wire resistance and test lead wire resistance, a very low ohmic value.
:
:Move the probe over to the right of R11 [Reference box B] and then be expecting in the order of ~22 ohms, and then move down to the 60 ufd electrolytic and confirm the same resistance at its pos terminal, then probe up to the right side of R10, above, for the same resistance.
:
:The sets main B+ buss then routes up to the power input side of the audio output which will probably not be terminal accessible, so move on past it and read for continuity at the plate ( left un-numbered by me, since I was unsure which AF out tube was being used.... pin 3 if it is a 50L6.
:
:That should read a cumulative ~180 + 22 ohms total.
:
:If all checks out, go back down to the R10 1000ohm voltage dropping / filter resistor which drops that main ~110Vdc down to a sub B+ level of ~ 93 V [ marked up in blue squares] and is started at [ Reference box C].
:
:It makes 4 branch offs to feed circuits, the first one would be a direct run up to the screen grid of the AF output tube.
:
:Take the negative ohmmeter lead and keep it on [ Reference box C] and use the other probe to check out [ Reference box C], where another low reading would be expected for a direct wire linking.
:
:Then you move the probe over to pin 5 of the 12BE6 [ Reference box F] and then to pin 5 of the 12BA6[ Reference box G] , where you should expect ~ 20 ohms resistance from IF transformers primary windings being read out in the series path of either tube.
:
:The last leg to be measured is going to encounter a 470K plate resistor in line, so when metering the plate of the 12AT6, expect ~470k resistance at [ Reference box H] . That should be enough initial continuity of B+ checks.
:
:
:Move the ohmmeters neg ground lead to the black test box marked Gnd and then take the probe and check for continuity over to the two black G box references that are extensions of the initial gnd buss...low ohmmages..right ? as that should only be wire interconnects....if nothing was disrupted in parts change outs.
:I really would suspect the trouble to be in the area of the filter capacitors wiring or the associated R10 and R11 resistors interconnecting wiring, where you might have made a miss wire or left something unconnected.
:
:As far as what all parts you consider being involved in a "recapping" of that set, I would be initially considering the filter condensers and then look at the green mark up boxes for the marked capacitors. The number one most critical , being the heavy marked up one at the 1st grid of the AF output tube ( unless Zenith used a disc ceramic type for it...as well as disc ceramics being used in other places also.)[That type is a very,very, very trouble free capacitor.]
:
:I also don't know if your unit used the little U1 couplate instead of the individual discrete res and caps between the 12AT6 and the output tube circuitry.
:
:
:If C3 is using a paper type of cap, it might be a 'BERRY good idea to replace it with a modern day MKT series of capacitor that is safety rated for use directly across an AC line. When a conventional paper cap is used there, all it takes is for a high spike voltage from the AC line / and-or / a lightning hits induced voltage coming in to present itself across the caps electrodes and a punch through of the dielectric . That initial plasma path then permits the whole 120 volt line voltage continually dump its fury across that conductive path and then there is fireworks until that path burns open enough or the cap explodes.
:
:
:
:If it seems that the power supply is operative with voltages, you then might try a testing of the power supply and the AF output stages by taking a test lead with clips at its ends and take a spare .001 thru -.1 ufd capacitor and place one test lead end clip on pin 4 of the 12AT6 to extract a 6 Vac test signal and take the other free end of the test lead and clip on to one end of that capacitor, with the other cap lead used to initially / sequentially touch to the 1st grid of the AF output tube…expecting hum if all is well…. and then move the free cap lead up to the 1st grid of the 12AT6 expecting much more hum and then finally turn the volume control to max and and move the cap over to inject signal into terminal 4&5 at the bottom of the secondary of the 2nd IF transformer. All of these are sequentially numbered on the markup as [1] [2] [3].
:
:
:
:Standing by for feed back..... as the next step might be powering up and checking for the tubes screen and plate voltages.
:
:
:
:
:
:73's de Edd
:
:
:
:
:
:
:
:
:
::::Hello out there,...can you hear me now ?
:::: I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
::::Thanks, Joe G.
:::
:::Joe,
:::Check for voltage at switch input and out. You could have a faulty switch contact.
:::Jon.
::::Jon, Switch has power in and out. Also picked up output for volume. Thanks,Joe

2/12/2008 8:30:21 PMJoe G
:
:
:
:
:
:Possibly, I can throw in a bit more.
:
:With the initial assumption that the tube filaments and the dial light are lighting up. Also, that the unit was working in a fashion before you changed any parts. That certainly leaves a miswire or a connection error most suspect.
:
:Since you mentioned schematicum unfamiliaritis , I am attaching a mark-up working schema, where you can just track down to its specified test points.
:
:
:
:Schema insert:
:
:
:
:
:
:
::Initially take a power down and do an ohmming out of the main B+ line [ marked up in red squares] which originates at the 35W4 pin 7 [Reference box A]and travels down to the left lead of surge inrush resistor R11...so ohms wise, you should only be reading wire resistance and test lead wire resistance, a very low ohmic value.
:
:Move the probe over to the right of R11 [Reference box B] and then be expecting in the order of ~22 ohms, and then move down to the 60 ufd electrolytic and confirm the same resistance at its pos terminal, then probe up to the right side of R10, above, for the same resistance.
:
:The sets main B+ buss then routes up to the power input side of the audio output which will probably not be terminal accessible, so move on past it and read for continuity at the plate ( left un-numbered by me, since I was unsure which AF out tube was being used.... pin 3 if it is a 50L6.
:
:That should read a cumulative ~180 + 22 ohms total.
:
:If all checks out, go back down to the R10 1000ohm voltage dropping / filter resistor which drops that main ~110Vdc down to a sub B+ level of ~ 93 V [ marked up in blue squares] and is started at [ Reference box C].
:
:It makes 4 branch offs to feed circuits, the first one would be a direct run up to the screen grid of the AF output tube.
:
:Take the negative ohmmeter lead and keep it on [ Reference box C] and use the other probe to check out [ Reference box C], where another low reading would be expected for a direct wire linking.
:
:Then you move the probe over to pin 5 of the 12BE6 [ Reference box F] and then to pin 5 of the 12BA6[ Reference box G] , where you should expect ~ 20 ohms resistance from IF transformers primary windings being read out in the series path of either tube.
:
:The last leg to be measured is going to encounter a 470K plate resistor in line, so when metering the plate of the 12AT6, expect ~470k resistance at [ Reference box H] . That should be enough initial continuity of B+ checks.
:
:
:Move the ohmmeters neg ground lead to the black test box marked Gnd and then take the probe and check for continuity over to the two black G box references that are extensions of the initial gnd buss...low ohmmages..right ? as that should only be wire interconnects....if nothing was disrupted in parts change outs.
:I really would suspect the trouble to be in the area of the filter capacitors wiring or the associated R10 and R11 resistors interconnecting wiring, where you might have made a miss wire or left something unconnected.
:
:As far as what all parts you consider being involved in a "recapping" of that set, I would be initially considering the filter condensers and then look at the green mark up boxes for the marked capacitors. The number one most critical , being the heavy marked up one at the 1st grid of the AF output tube ( unless Zenith used a disc ceramic type for it...as well as disc ceramics being used in other places also.)[That type is a very,very, very trouble free capacitor.]
:
:I also don't know if your unit used the little U1 couplate instead of the individual discrete res and caps between the 12AT6 and the output tube circuitry.
:
:
:If C3 is using a paper type of cap, it might be a 'BERRY good idea to replace it with a modern day MKT series of capacitor that is safety rated for use directly across an AC line. When a conventional paper cap is used there, all it takes is for a high spike voltage from the AC line / and-or / a lightning hits induced voltage coming in to present itself across the caps electrodes and a punch through of the dielectric . That initial plasma path then permits the whole 120 volt line voltage continually dump its fury across that conductive path and then there is fireworks until that path burns open enough or the cap explodes.
:
:
:
:If it seems that the power supply is operative with voltages, you then might try a testing of the power supply and the AF output stages by taking a test lead with clips at its ends and take a spare .001 thru -.1 ufd capacitor and place one test lead end clip on pin 4 of the 12AT6 to extract a 6 Vac test signal and take the other free end of the test lead and clip on to one end of that capacitor, with the other cap lead used to initially / sequentially touch to the 1st grid of the AF output tube…expecting hum if all is well…. and then move the free cap lead up to the 1st grid of the 12AT6 expecting much more hum and then finally turn the volume control to max and and move the cap over to inject signal into terminal 4&5 at the bottom of the secondary of the 2nd IF transformer. All of these are sequentially numbered on the markup as [1] [2] [3].
:
:
:
:Standing by for feed back..... as the next step might be powering up and checking for the tubes screen and plate voltages.
:
:
:
:
:
:73's de EddEdd,
Thank you so much for taking the time and putting this all down for me.
:I probably wont have the time to get to it until this weekend. I’ve been working a lot of hours lately. The tubes do not light up though and neither does the pilot light. I tested the tubes today one took along time to slowly climb up to the point where it regitsered good, but all the others checked fine. Thanks again, Joe.
:
:
:
:
:
:
:
:
::::Hello out there,...can you hear me now ?
:::: I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
::::Thanks, Joe G.
:::
:::Joe,
:::Check for voltage at switch input and out. You could have a faulty switch contact.
:::Jon.
::::Jon, Switch has power in and out. Also picked up output for volume. Thanks,Joe
2/12/2008 8:42:54 PMdel in mn
This project reminds me of the heart surgeon or dentist who said 'oops'.
2/13/2008 11:32:05 AMEdd




Looks like you need info on the filament string also then, not sure if your set incorporated the 50L6 instead, so its filament mark up is just below the 50C5…with the 2 and 7 pins possibly being transposed in the sets wiring sequence.

The filament string is the orange path, if you would initially take an ohmmeter lead and stab it into pin 3 of the 12AT6 you could then take that other lead and check for very low path resistance to chassis ground and then if it was moved to one of the two pins on the AC cord there should be a like low resistance reading with the volume controls power switch turned on. If so, flip the switch on and off several times to confirm its integrity.
That confirms continuity from the AC plugs pin all the way thru to pin 3 of the 12AT6.

By leaving that ohmmeter connection on the AC plug , you then could walk the probe that was initially on 12AT6 pin 3 over to its pin 4 and expect ~12 -15 ohms to be in line, then walk up the series path to12BE6 with its additional ~12 -15 ohms resistance showing up, then the12BA6 with its additional ~12 -15 ohms and then to the 50C5 (50L6?) which might tack on an additional ~30 ohms for it alone.
That should bring you right to the last tube….pin 3 of the 35W4 and thru it and coming out its pin 4 with an additional ~30 ohms added to the strings resistance. The whole strings resistance is somewhat dependent upon the type of metering used and the actual voltage presented in its measurement. There will be a slight shift downwards in resistance as the string tries to heat up but a mere bit with that slight voltage being applied. Overall you might expect anywhere from ~180ohms on downward to ~120 ohms for the whole strings series resistance from [ F ] thru [ F’ ].

You’re NOT expecting an open circuit…. which you apparently have now.

By this technique you have confirmed ALL continuity from pin to pin of the AC cord thru the switch and the filament string, with only a “pulling apart” and flexing of the AC cord across its length to confirm ITS integrity….(but this is not a cord on an electric heating iron …or a soldering irons cord….where that might be suspect from intensive flexure.)


New markup with its additions is just below:


73's de Edd






::
::
::
::
::
::Possibly, I can throw in a bit more.
::
::With the initial assumption that the tube filaments and the dial light are lighting up. Also, that the unit was working in a fashion before you changed any parts. That certainly leaves a miswire or a connection error most suspect.
::
::Since you mentioned schematicum unfamiliaritis , I am attaching a mark-up working schema, where you can just track down to its specified test points.
::
::
::
::Schema insert:
::
::
::
::
::
::
::::Initially take a power down and do an ohmming out of the main B+ line [ marked up in red squares] which originates at the 35W4 pin 7 [Reference box A]and travels down to the left lead of surge inrush resistor R11...so ohms wise, you should only be reading wire resistance and test lead wire resistance, a very low ohmic value.
::
::Move the probe over to the right of R11 [Reference box B] and then be expecting in the order of ~22 ohms, and then move down to the 60 ufd electrolytic and confirm the same resistance at its pos terminal, then probe up to the right side of R10, above, for the same resistance.
::
::The sets main B+ buss then routes up to the power input side of the audio output which will probably not be terminal accessible, so move on past it and read for continuity at the plate ( left un-numbered by me, since I was unsure which AF out tube was being used.... pin 3 if it is a 50L6.
::
::That should read a cumulative ~180 + 22 ohms total.
::
::If all checks out, go back down to the R10 1000ohm voltage dropping / filter resistor which drops that main ~110Vdc down to a sub B+ level of ~ 93 V [ marked up in blue squares] and is started at [ Reference box C].
::
::It makes 4 branch offs to feed circuits, the first one would be a direct run up to the screen grid of the AF output tube.
::
::Take the negative ohmmeter lead and keep it on [ Reference box C] and use the other probe to check out [ Reference box C], where another low reading would be expected for a direct wire linking.
::
::Then you move the probe over to pin 5 of the 12BE6 [ Reference box F] and then to pin 5 of the 12BA6[ Reference box G] , where you should expect ~ 20 ohms resistance from IF transformers primary windings being read out in the series path of either tube.
::
::The last leg to be measured is going to encounter a 470K plate resistor in line, so when metering the plate of the 12AT6, expect ~470k resistance at [ Reference box H] . That should be enough initial continuity of B+ checks.
::
::
::Move the ohmmeters neg ground lead to the black test box marked Gnd and then take the probe and check for continuity over to the two black G box references that are extensions of the initial gnd buss...low ohmmages..right ? as that should only be wire interconnects....if nothing was disrupted in parts change outs.
::I really would suspect the trouble to be in the area of the filter capacitors wiring or the associated R10 and R11 resistors interconnecting wiring, where you might have made a miss wire or left something unconnected.
::
::As far as what all parts you consider being involved in a "recapping" of that set, I would be initially considering the filter condensers and then look at the green mark up boxes for the marked capacitors. The number one most critical , being the heavy marked up one at the 1st grid of the AF output tube ( unless Zenith used a disc ceramic type for it...as well as disc ceramics being used in other places also.)[That type is a very,very, very trouble free capacitor.]
::
::I also don't know if your unit used the little U1 couplate instead of the individual discrete res and caps between the 12AT6 and the output tube circuitry.
::
::
::If C3 is using a paper type of cap, it might be a 'BERRY good idea to replace it with a modern day MKT series of capacitor that is safety rated for use directly across an AC line. When a conventional paper cap is used there, all it takes is for a high spike voltage from the AC line / and-or / a lightning hits induced voltage coming in to present itself across the caps electrodes and a punch through of the dielectric . That initial plasma path then permits the whole 120 volt line voltage continually dump its fury across that conductive path and then there is fireworks until that path burns open enough or the cap explodes.
::
::
::
::If it seems that the power supply is operative with voltages, you then might try a testing of the power supply and the AF output stages by taking a test lead with clips at its ends and take a spare .001 thru -.1 ufd capacitor and place one test lead end clip on pin 4 of the 12AT6 to extract a 6 Vac test signal and take the other free end of the test lead and clip on to one end of that capacitor, with the other cap lead used to initially / sequentially touch to the 1st grid of the AF output tube…expecting hum if all is well…. and then move the free cap lead up to the 1st grid of the 12AT6 expecting much more hum and then finally turn the volume control to max and and move the cap over to inject signal into terminal 4&5 at the bottom of the secondary of the 2nd IF transformer. All of these are sequentially numbered on the markup as [1] [2] [3].
::
::
::
::Standing by for feed back..... as the next step might be powering up and checking for the tubes screen and plate voltages.
::
::
::
::
::
::73's de EddEdd,
: Thank you so much for taking the time and putting this all down for me.
::I probably wont have the time to get to it until this weekend. I’ve been working a lot of hours lately. The tubes do not light up though and neither does the pilot light. I tested the tubes today one took along time to slowly climb up to the point where it regitsered good, but all the others checked fine. Thanks again, Joe.
::
::
::
::
::
::
::
::
:::::Hello out there,...can you hear me now ?
::::: I have a Zenith H511G that was working until I recapped it. I recapped two other radios and replaced the dial cords with no problems. The one a Crosley didn’t work before but does now. I’ve gone over all my solder points and they look O.K. I have no power that I can detect. I have 117 Volts at the power cord ends. I replaced the 50L6GT tube, still nothing. I don’t read schematics very well. What can I do and where do I start.
:::::Thanks, Joe G.
::::
::::Joe,
::::Check for voltage at switch input and out. You could have a faulty switch contact.
::::Jon.
:::::Jon, Switch has power in and out. Also picked up output for volume. Thanks,Joe



© 1989-2025, Nostalgia Air