813 Based Amps - UK Vintage Radio Repair and Restoration Discussion Forum

Skywave · 3rd Jan 2007, 11:49 pm

[Split from this thread :-

https://www.vintage-radio.net/forum/...ad.php?t=13737
]

Quote:

Originally Posted by boiss

Valves are 813s . . . . I like the idea of swapping the valves every now and then. Simple is what works best!
Happy New Year,
Mike.

Just a thought . . .
Are you planning on a pair of 813s in push-pull - and possibly with neutralisation on each? If so, the idea of swapping bottles may cause problems, since the neutralising caps. will need re-setting when the bottles are swapped over.

Even if you will be using them in parallel, this is possibly still a concern - depending on the cct. arrangements.

A further thought on heaters for 813s just entered my mind: since these valves are rather hungry heater current-wise, (10v @ 5amps each) make sure that you have the correct voltage when measured at the valve pins - needs to be within +/- 5%. Insufficient heater voltage is a common cause of premature valve failure in bottles of this type

.

Just a couple of thoughts - good luck with it -

Al.

boiss · 5th Jan 2007, 12:48 pm

This is an Audio amp not RF, I hope. As for the correct heater volts there is a feedback cct to monitor the volts at the load (heater). On the Low volts theme, there are many conflicting opinions, early writings were very much along the lines of too much or too low a voltage will seriously shorten valve life but other more recent opinions are that reducing heater volts (-5% to -10%) prolongs valve life. I will set mine at 10V anyway. I have been using a very low power 813 modulator with 8V on the heater and Va 25V, the 813 was second hand, over the last few months it has clocked up a few hundred hrs and there doesn't appear to have been any change in performance so far. I remember reading somewhere that the thoriated tungstem heater could be poisoned if left on with no Ht and that generally the heater could be restored to normal by applying 12V for around 15 mins.
Mike.

Skywave · 5th Jan 2007, 7:36 pm

Quote:

Originally Posted by boiss

This is an Audio amp not RF, I hope. Mike.

I see.

But I'm sure you can understand my assumption that you were aiming at an RF amp. - since an 813 was designed primarily for that type of service.

Good luck with it anyway.

/ Al.

Alan_Douglas · 7th Jan 2007, 10:15 pm

In general, if the filament voltage is too low, the space charge is insufficient to neutralize positive ions which can then hit the cathode and poison it.

cathoderesistor · 8th Jan 2007, 12:11 pm

Hello Mike,
Ah! 813, makes perfect sense now.

If you ever fancy making a PP one

http://www.izzy-wizzy.com/audio/powamp813.html

I dare say you already know about Steve Bench's and Pete Millet's 813 amps.

boiss · 8th Jan 2007, 1:29 pm

Hi CR, It will be PP, I've already seen these designs and the one in the link with all the transformers would be hideously expensive. I have given the problems a little bit of thought and will probably use cathode follower drivers maybe EL84s strapped as triodes although it does seem a bit of overkill. ECC8xs would probably be hard pushed to drive enough volts into around 15K.
Once I get my O/P tranny finished I will start a new post.
Mike.

Ed_Dinning · 8th Jan 2007, 8:51 pm

Hi Mike, the GEC book "An Approach to Audio Frequency Amplifier Design" has a hodt of curcuits up to 1100W! It also gives plenty of design hints on driving the output valves. PM me if you need more details.

Ed

oldeurope · 9th Jan 2007, 9:47 am

Quote:

Originally Posted by cathoderesistor

Hello Mike,
Ah! 813, makes perfect sense now.

If you ever fancy making a PP one

http://www.izzy-wizzy.com/audio/powamp813.html

I dare say you already know about Steve Bench's and Pete Millet's 813 amps.

Hm, I don't have a good feeling with unscreened valves at high (means more than 500V) plate voltages at high (more than 20mA) currents. X-ray below 5KeV can be screened easiely.

Kind regards,
Darius

boiss · 10th Jan 2007, 6:39 pm

I would imagine that Xrays produced by thermionic valves using around 1KV would be extremely soft and not go through the anode structure and then the glass envelope, (Leaded?). In the past when I worked for Marconi it was only necessary to wear a film badge when working on high voltage equipments, generally 15KV plus. Tv Line stages in BW sets often had 800V pk on the valves and some simple colour sets had 1500V without any screening. My 1938 Telefunken radio has 480V on the anode.
Is there really more danger in sitting 6 feet away from an amplifier with 1KV on it's valves or 6ft in front of a TV ?
While I accept that there may be some X radiation how does it compare to having a couple of Xrays done at the dentist every year, and at 1KeV what is the penetration wrt the human body clothing etc?
Mike.

oldeurope · 10th Jan 2007, 7:06 pm

My 1938 Telefunken radio has 480V on the anode.

Yes, this is below 500V. No problem. How much volts does the 813 have there? Not that there must be plate current and plate voltage to produce x-ray!!!

Darius

boiss · 10th Jan 2007, 8:33 pm

Anode volts between 800V and 1KV and Ia somewhere between 60-100 mA, however the wavelength and thence penetration properties are wavelength dependent aren't they?
Mike

boiss · 1st Feb 2007, 12:22 am

Ok All, Next problem to solve. How to monitor Ia of each valve so that the DC state of each valve can be balanced by varying the g1 bias. (Both heaters share the same SMPS). The idea of trying to measure Ia at the anodes and having about 1KV on the switch and the meter doesn't appear very safe.
Mike.

ppppenguin · 1st Feb 2007, 9:00 am

Why not measure at the cathodes. I know this includes the screen grid current but that should be a fairly small fraction of the anode current.

boiss · 1st Feb 2007, 3:19 pm

Hi Jeffrey,
Because the 813s are directly heated, ie. the heater is the cathode, and in this case the heaters are connected together because they share the same power supply.
Mike.

Colin · 1st Feb 2007, 3:28 pm

Mike,

This idea may be worth investigating ...

http://home.cogeco.ca/~rpaisley4/DCCAmmeter.html

I haven't looked into it in any detail so it may not be suitable for the voltages and currents you're working with - but worth a look just in case.

rgds
Colin

Denis G4DWC · 1st Feb 2007, 5:24 pm

If it were me I would investigate the use of LEM Hall effect Current transducers. You could measure the cathode current by passing both heater wires through the core. The Magnetic field due to heater current would cancel and you would be left with the current due to anode and screen current. A simpler way would be to use a LEM device to measure Anode current directly. Some of the small LEM modules have 3 loops through the core moulded in so you could put them in series to increase the sensitivity. The insulation is around 3kV. Alternatively you could use a tranducer with a hole and feed the HT lead several times through it. Although they appear to have high current ratings they are accurate at low currents as well. Farnell do lots of different LEM current transducers that are not too expensive. Just search for LEM and you'll turn up about 50 different ones. I have used quite a few of them in different designs. There is loads of info on the LEM.com website you may find useful......HTH
Denis

GMB · 1st Feb 2007, 6:52 pm

The vintage technology solution to measuring current like this is the HP428B if you can find one. It uses a clip-on fluxgate probe and can even measure quite small DC currents.

Skywave · 1st Feb 2007, 8:11 pm

Quote:

Originally Posted by boiss

Ok All, Next problem to solve. How to monitor Ia of each valve so that the DC state of each valve can be balanced by varying the g1 bias. Mike.

Ok - I know that what I am about to suggest is a little unconventional, and I will talk only in concept terms.

Insert a small current-sensing resistor in each anode supply line. Use the voltage developed across this to drive a one transistor dc amplifier with a sensitive mA meter in the collector. Run it from a small 9v battery. The current will be very small - the battery should last ages - especially if you fit a "push-to-read" switch in series with the meter. Fit the whole affair - including meter - into a well-insulated box; meter on the front panel. Do this twice - once for each valve.

Just an 'off-the-cuff' thought.

Sometimes awkward problems require unusual approaches to find thier solution.

Al / Skywave.

boiss · 1st Feb 2007, 8:25 pm

Thanks for the ideas, I think the LEM one looks interesting, the sensing resistor was the simple choice but I still don't like the idea of having HT in different places under and around the chassis. Had a thought, how about a way to measure flux in the core of the tranny, that would mean an absolute method of adjustment.
Mike.

boiss · 1st Feb 2007, 11:11 pm

OK Here is a really silly idea, or is it??
The result of balancing the Valve currents, Ia and Ig2 is to reduce to a minimum the standing magnetizing current ln the transformer. In the case of a toroidal tranny without an airgap this is more important than ever but the flux path is fairly well defined around the core rather than the split in flux density around an EI core.
And now for the silly bit.
If a small compass was fixed to the core it would indicate the flux direction and the null point would tend to be transverse to the core. Would it be sensitive enough? Could it work?
Any thoughts on this or are you going to call for them in the white coats to come and get me!!
Mike

5th Jan 2007, 12:48 pm	#2
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: Directly heated cathodes This is an Audio amp not RF, I hope. As for the correct heater volts there is a feedback cct to monitor the volts at the load (heater). On the Low volts theme, there are many conflicting opinions, early writings were very much along the lines of too much or too low a voltage will seriously shorten valve life but other more recent opinions are that reducing heater volts (-5% to -10%) prolongs valve life. I will set mine at 10V anyway. I have been using a very low power 813 modulator with 8V on the heater and Va 25V, the 813 was second hand, over the last few months it has clocked up a few hundred hrs and there doesn't appear to have been any change in performance so far. I remember reading somewhere that the thoriated tungstem heater could be poisoned if left on with no Ht and that generally the heater could be restored to normal by applying 12V for around 15 mins. Mike.

7th Jan 2007, 10:15 pm	#4
Alan_Douglas Retired Dormant Member Join Date: Nov 2003 Location: Cape Cod, Massachusetts Posts: 59	Re: Directly heated cathodes In general, if the filament voltage is too low, the space charge is insufficient to neutralize positive ions which can then hit the cathode and poison it.

8th Jan 2007, 12:11 pm	#5
cathoderesistor Retired Dormant Member Join Date: Dec 2006 Location: London N1 Posts: 3	Re: Directly heated cathodes Hello Mike, Ah! 813, makes perfect sense now. If you ever fancy making a PP one http://www.izzy-wizzy.com/audio/powamp813.html I dare say you already know about Steve Bench's and Pete Millet's 813 amps.

8th Jan 2007, 1:29 pm	#6
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: Directly heated cathodes Hi CR, It will be PP, I've already seen these designs and the one in the link with all the transformers would be hideously expensive. I have given the problems a little bit of thought and will probably use cathode follower drivers maybe EL84s strapped as triodes although it does seem a bit of overkill. ECC8xs would probably be hard pushed to drive enough volts into around 15K. Once I get my O/P tranny finished I will start a new post. Mike.

8th Jan 2007, 8:51 pm	#7
Ed_Dinning Dekatron Join Date: Jan 2004 Location: Newcastle upon Tyne, Tyne & Wear, UK. Posts: 8,194	Re: 813 Based Amps Hi Mike, the GEC book "An Approach to Audio Frequency Amplifier Design" has a hodt of curcuits up to 1100W! It also gives plenty of design hints on driving the output valves. PM me if you need more details. Ed

10th Jan 2007, 6:39 pm	#9
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: 813 Based Amps I would imagine that Xrays produced by thermionic valves using around 1KV would be extremely soft and not go through the anode structure and then the glass envelope, (Leaded?). In the past when I worked for Marconi it was only necessary to wear a film badge when working on high voltage equipments, generally 15KV plus. Tv Line stages in BW sets often had 800V pk on the valves and some simple colour sets had 1500V without any screening. My 1938 Telefunken radio has 480V on the anode. Is there really more danger in sitting 6 feet away from an amplifier with 1KV on it's valves or 6ft in front of a TV ? While I accept that there may be some X radiation how does it compare to having a couple of Xrays done at the dentist every year, and at 1KeV what is the penetration wrt the human body clothing etc? Mike.

10th Jan 2007, 7:06 pm	#10
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	x-ray My 1938 Telefunken radio has 480V on the anode. Yes, this is below 500V. No problem. How much volts does the 813 have there? Not that there must be plate current and plate voltage to produce x-ray!!! Darius

10th Jan 2007, 8:33 pm	#11
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: 813 Based Amps Anode volts between 800V and 1KV and Ia somewhere between 60-100 mA, however the wavelength and thence penetration properties are wavelength dependent aren't they? Mike

1st Feb 2007, 12:22 am	#12
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: 813 Based Amps Ok All, Next problem to solve. How to monitor Ia of each valve so that the DC state of each valve can be balanced by varying the g1 bias. (Both heaters share the same SMPS). The idea of trying to measure Ia at the anodes and having about 1KV on the switch and the meter doesn't appear very safe. Mike.

1st Feb 2007, 9:00 am	#13
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: 813 Based Amps Why not measure at the cathodes. I know this includes the screen grid current but that should be a fairly small fraction of the anode current.

1st Feb 2007, 3:19 pm	#14
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: 813 Based Amps Hi Jeffrey, Because the 813s are directly heated, ie. the heater is the cathode, and in this case the heaters are connected together because they share the same power supply. Mike.

1st Feb 2007, 3:28 pm	#15
Colin Heptode Join Date: Jun 2004 Location: Ilkley, West Yorkshire, UK. Posts: 656	Re: 813 Based Amps Mike, This idea may be worth investigating ... http://home.cogeco.ca/~rpaisley4/DCCAmmeter.html I haven't looked into it in any detail so it may not be suitable for the voltages and currents you're working with - but worth a look just in case. rgds Colin

1st Feb 2007, 5:24 pm	#16
Denis G4DWC Pentode Join Date: Oct 2006 Location: Brackley Northamptonshire, UK. Posts: 240	Re: 813 Based Amps If it were me I would investigate the use of LEM Hall effect Current transducers. You could measure the cathode current by passing both heater wires through the core. The Magnetic field due to heater current would cancel and you would be left with the current due to anode and screen current. A simpler way would be to use a LEM device to measure Anode current directly. Some of the small LEM modules have 3 loops through the core moulded in so you could put them in series to increase the sensitivity. The insulation is around 3kV. Alternatively you could use a tranducer with a hole and feed the HT lead several times through it. Although they appear to have high current ratings they are accurate at low currents as well. Farnell do lots of different LEM current transducers that are not too expensive. Just search for LEM and you'll turn up about 50 different ones. I have used quite a few of them in different designs. There is loads of info on the LEM.com website you may find useful......HTH Denis __________________ If you take something apart often enough, eventually you'll have two of them....

1st Feb 2007, 6:52 pm	#17
GMB Dekatron Join Date: Aug 2003 Location: near Reading (and sometimes Torquay) Posts: 3,095	Re: 813 Based Amps The vintage technology solution to measuring current like this is the HP428B if you can find one. It uses a clip-on fluxgate probe and can even measure quite small DC currents.

1st Feb 2007, 8:25 pm	#19
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: 813 Based Amps Thanks for the ideas, I think the LEM one looks interesting, the sensing resistor was the simple choice but I still don't like the idea of having HT in different places under and around the chassis. Had a thought, how about a way to measure flux in the core of the tranny, that would mean an absolute method of adjustment. Mike.

1st Feb 2007, 11:11 pm	#20
boiss Pentode Join Date: Jun 2006 Location: Malaga, Spain. Posts: 235	Re: 813 Based Amps OK Here is a really silly idea, or is it?? The result of balancing the Valve currents, Ia and Ig2 is to reduce to a minimum the standing magnetizing current ln the transformer. In the case of a toroidal tranny without an airgap this is more important than ever but the flux path is fairly well defined around the core rather than the split in flux density around an EI core. And now for the silly bit. If a small compass was fixed to the core it would indicate the flux direction and the null point would tend to be transverse to the core. Would it be sensitive enough? Could it work? Any thoughts on this or are you going to call for them in the white coats to come and get me!! Mike