Static Gm test

PeteMcK · 26th February 2009, 11:06 AM

Will the attached work as a static Gm test?
If Ip1 = Vk1/Rk1 and Ip2 = Vk2/(Rk1+Rk2),
then gm = (Ip1-Ip2)/(Vk1-Vk2)...???

Brian Beck · 26th February 2009, 04:05 PM

Yes, this should give you gm. I would suggest making Rk2 a small percentage of Rk1 in your schematic. Set up Rk1 to give you the desired operating point for the tube - the point where you want to know its gm. That way the change in Vgk and Ip is a small percentage. Gm is a small signal (linear assumption) parameter, dIp/dVgk. With a typical DVM, accuracy is good enough to extract gm to a high degree of accuracy with very small changes in Vgk and Ip.

PeteMcK · 26th February 2009, 08:35 PM

Thanks Brian, good to know I'm on the right track.

(I have 30 or 40 12ax7s (mostly Philips) I pulled out of an old church organ over 30 years ago, so I just want a simple way to test them)

Cheers,
Pete McK

Brian Beck · 26th February 2009, 10:12 PM

To be even more precise, you'd really want to hold Vpk constant when measuring gm. In your case Vpk moves by the same amount as Vgk, but plate current changes due to Vpk are mu times less sensitive than plate current changes due to Vgk. For medium to high mu tubes, this technique will do well enough. For low mu tubes, you'll run into mu/(mu+1) errors. A better technique, perhaps not quite as simple, is to ground the cathode via a small sensing resistor, say 1 to 10 ohms, and set a HV power supply at the desired Vpk on the plate. Apply an adjustable low-voltage negative supply on the grid. You can step the grid voltage and monitor plate current changes on a DVM connected across the small current sensing resistor. You can also vary the plate voltages and recreate the entire set of tube curves! Just make sure you don't exceed rated tube voltage, current and power values.

By the way, there's nothing wrong with these "static" tests. The results will apply at virtually all audio frequencies. Tube curves were made this way, after all.

PeteMcK · 27th February 2009, 04:03 AM

Thanks for the info Brian, although I don't really want to be that precise (but maybe one day... & I'm sure someone else will find it useful. It is useful in helping me re-learn stuff I'd forgotten from decades ago too).

I'm thinking guitar amp, so I'll probably use V+ 150-200V, Rp = 100K, Rg = 1M, Rk1 = 1K5, and maybe make Rk2 a low value preset, so I can play with it a bit...
cheers,
Pete McK

Brian Beck · 27th February 2009, 02:31 PM

Oops, I see now that you show a plate resistor. I would eliminate it and replace it with a supply. Sorry, my arguments even for your approach with resistor switching in the cathode assumed no plate resistor. I just didn't pay enough attention to your first image. The plate will move down enough to significantly change Vpk and upset your reading. Vpk must be held nearly constant during gm tests. You may get by with your technique for doing a relative check of the health of your tubes. If you observe how a known-good tube checks, you can estimate whether the others are good or bad even if you can't arrive at a good gm number.

PeteMcK · 2nd March 2009, 07:05 PM

OK Brian, a fixed supply should be easy enough, but then I suppose the issue becomes how to limit the current through the valve?
or alternatively, an adjustable supply with the Vp being made the same for both readings ?

26th February 2009, 04:05 PM	#2
Brian Beck diyAudio Member Join Date: Dec 2004 Location: South Florida, USA	Yes, this should give you gm. I would suggest making Rk2 a small percentage of Rk1 in your schematic. Set up Rk1 to give you the desired operating point for the tube - the point where you want to know its gm. That way the change in Vgk and Ip is a small percentage. Gm is a small signal (linear assumption) parameter, dIp/dVgk. With a typical DVM, accuracy is good enough to extract gm to a high degree of accuracy with very small changes in Vgk and Ip. __________________ Brian

26th February 2009, 08:35 PM	#3
PeteMcK diyAudio Member Join Date: Feb 2002 Location: Sydney	Thanks Brian, good to know I'm on the right track. (I have 30 or 40 12ax7s (mostly Philips) I pulled out of an old church organ over 30 years ago, so I just want a simple way to test them) Cheers, Pete McK __________________ ‘today… there lives alongside the twentieth century the tenth or thirteenth. A hundred million people use electricity and still believe in the magic power of signs and exorcisms” Trotsky

26th February 2009, 10:12 PM	#4
Brian Beck diyAudio Member Join Date: Dec 2004 Location: South Florida, USA	To be even more precise, you'd really want to hold Vpk constant when measuring gm. In your case Vpk moves by the same amount as Vgk, but plate current changes due to Vpk are mu times less sensitive than plate current changes due to Vgk. For medium to high mu tubes, this technique will do well enough. For low mu tubes, you'll run into mu/(mu+1) errors. A better technique, perhaps not quite as simple, is to ground the cathode via a small sensing resistor, say 1 to 10 ohms, and set a HV power supply at the desired Vpk on the plate. Apply an adjustable low-voltage negative supply on the grid. You can step the grid voltage and monitor plate current changes on a DVM connected across the small current sensing resistor. You can also vary the plate voltages and recreate the entire set of tube curves! Just make sure you don't exceed rated tube voltage, current and power values. By the way, there's nothing wrong with these "static" tests. The results will apply at virtually all audio frequencies. Tube curves were made this way, after all. __________________ Brian

27th February 2009, 04:03 AM	#5
PeteMcK diyAudio Member Join Date: Feb 2002 Location: Sydney	Thanks for the info Brian, although I don't really want to be that precise (but maybe one day... & I'm sure someone else will find it useful. It is useful in helping me re-learn stuff I'd forgotten from decades ago too). I'm thinking guitar amp, so I'll probably use V+ 150-200V, Rp = 100K, Rg = 1M, Rk1 = 1K5, and maybe make Rk2 a low value preset, so I can play with it a bit... cheers, Pete McK __________________ ‘today… there lives alongside the twentieth century the tenth or thirteenth. A hundred million people use electricity and still believe in the magic power of signs and exorcisms” Trotsky

27th February 2009, 02:31 PM	#6
Brian Beck diyAudio Member Join Date: Dec 2004 Location: South Florida, USA	Oops, I see now that you show a plate resistor. I would eliminate it and replace it with a supply. Sorry, my arguments even for your approach with resistor switching in the cathode assumed no plate resistor. I just didn't pay enough attention to your first image. The plate will move down enough to significantly change Vpk and upset your reading. Vpk must be held nearly constant during gm tests. You may get by with your technique for doing a relative check of the health of your tubes. If you observe how a known-good tube checks, you can estimate whether the others are good or bad even if you can't arrive at a good gm number. __________________ Brian

2nd March 2009, 07:05 PM	#7
PeteMcK diyAudio Member Join Date: Feb 2002 Location: Sydney	OK Brian, a fixed supply should be easy enough, but then I suppose the issue becomes how to limit the current through the valve? or alternatively, an adjustable supply with the Vp being made the same for both readings ? __________________ ‘today… there lives alongside the twentieth century the tenth or thirteenth. A hundred million people use electricity and still believe in the magic power of signs and exorcisms” Trotsky

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Cracking/Static in Pre-Amp	TheVille	Solid State	2	17th April 2009 07:42 PM
static removal from lp	dhole	Analogue Source	3	8th April 2005 08:22 PM
Speaker Static	Megahurt	Car Audio	4	25th March 2005 07:40 PM
Crackling/Static on my GC.	azira	Chip Amps	2	25th May 2004 09:48 PM
static	offroadbum	Solid State	1	4th February 2004 12:50 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read