Pentode amp experiment. - UK Vintage Radio Repair and Restoration Discussion Forum

YT2095UK · 27th Sep 2009, 5:06 pm

Inspired by Pauls post here: https://www.vintage-radio.net/forum/...81&postcount=2
stating that a Pentode could be used if wired/strapped as a triode, I tried this with a speaker on a home-made Valve Electronics set (like the old Science Fair sets but with Valve holders to the spring terminals) and decided to experiment a little.
Firstly I took Pauls advice and wired the pentode as described, it worked but was pulling a LOT of power through even without an input, I use a small 250v 10W bulb in line with the HT supply as an indicator when I`m messing about

so I tried all sorts of different configurations with these 3 grids to plays with, all connected togeter as one, closest to anode goes to anode and same with cathode, standard config and trying each as a Control grid etc...
Many Many combos!
and one I found that Does work and pulls very little, is g3 to anode, g2 to Cathode, and the g1 still as the control grid, although strapping g2 and g1 together makes little difference.

I connected my Longwire to the control grid (g1) and checked for RF, sure enough it pinned the needle when it touched the anode connection!

it seems to make a fairly reasonable RF amp as the longwire alone only moves the needle a little, all other configs do very little, the Above mentioned works great, and as I mentioned that g2 makes little difference straped to g1, for Audio this is the case but you get a little more RF taking g2 to cathode.

what I`m acually hearing out of the speakers is the same as when you connect crystal set headphones to Ant and Gnd, only louder

and NOT the same as connecting the same to the input of an audio amp no matter what the gain (I tried that too).

I`m using an EL34, a 503 (0.05uF) antenna coupling cap to the control grid and pulling that down to deck with a 3.9M Ohm resistor.
anode O/P is to the Mains in side of an 8VAC adaptor transformer from an old BT answering machine, feeding a 3 Ohm speaker.
although I doubt any of that is Too Critical appart from using a high ohm resistor.

does anyone know if this is a Normal/Known config for a pentode application?

Woohoo! I finaly managed to figure out how to UL pictures too, here is my Valve Proto-typing set.

paulsherwin · 27th Sep 2009, 5:28 pm

Are you sure you've identified the pins properly? G2 (pin 4) should go to the anode and G3 (pin 1) to the cathode in a triode wired EL34.

An EL34 is an odd valve to use in experiments like this since it is a very high power output valve. Something like an EF80 or, if you prefer octal valves, an EF36/37/37A/39 would work just as well and use much less power. I'm also unsure how an EL34 would perform at radio frequencies.

It's also unclear how you're applying bias to the valve.

Paul

YT2095UK · 27th Sep 2009, 5:51 pm

absolutely, pin for pin and double checked as always before I powered up.
in fact your wiring is excatly the same as I used in the OP stage of my amp that I made a few weeks ago, only I used a 1K resistor to take g2 to Anode rather than directly wiring it.
so I know what you said works

as for Biasing, there is none, the Cathode is Hard to deck, no RC coupling at all.
the Control grid (g1) is pulled down to deck via a 3.9M ohm, and fed with a 0.05uF cap.

I`m using the EL34 because I know that worst comes to worst, it wouldn`t matter too much How I wired it, the power I`m using wouldn`t destroy it without me knowing about it in plenty of time before hand.

as for RF on an EL34, I can say that LW is easily amplifiable (is that even a word

), I`ll have to get the sig-gen out and try that, noting the drop off (I`m keeping plenty of notes!).

Herald1360 · 27th Sep 2009, 8:44 pm

It's not surprising the EL34 takes a lot of current with zero bias!

As a triode with 250V on the anode/g2 and -15.5V on the grid (g1) it is supposed to take 70mA. With zero bias it goes off the data curves somewhere in excess of 250mA! To wind it back even to 10mA would need nearly -30V on the grid or somewhere around a 3.3K cathode resistor (bypassed for RF by say 100nF).

As an RF amplifier, in triode form it may well be unstable, if it's got any useful gain. Using it as a pentode would seem a better bet.....

Chris

YT2095UK · 27th Sep 2009, 9:16 pm

it did take a lot, but the way I have it configured currently, it takes very little (the bulb doesn`t even show except a little in IR.
g1 is the control as normal, g2 is direct to cathode, g3 is direct to anode.
set up like This, I get a really good RF amplification (micro watts to milli watts) and that`s whilst it`s driving a speaker.

now I realise that this in itself it`s practical for anything in particular as it stands, or with these particular parts I`v chosen to use, but I think there May be something useful here for Future work/experiments.

incidentaly, I did try a coupling to the sig-gen and it`s ok up until about 3MHz and totally drops off at 3.5.
I`ll replace the indutive load (OP transformer) with a resistor in the morning and try again, maybe the load plays a part?

Mr Moose · 28th Sep 2009, 12:05 pm

Hello,
With g3 connected to the anode you are running the valve as a tetrode rather than a pentode, although with g2 connected to the cathode it will act as a suppressor grid as well as a screen grid.
As Chris has said the anode current of an EL34 at zero bias and Va =250V, is over 250mA, however by connecting g2 to 0v you are probably reducing this to well below 25mA.
By connecting g1 to the cathode with a 3.9meg resistor you have a grid leak detector, although 3.9 meg is a higher than the normal 1meg to 2meg for a grid leak, you might well be getting AF as well as any residual RF at the output.
Yours, Richard

YT2095UK · 28th Sep 2009, 6:05 pm

I`m really glad you said that! I`d completely forgotten about Tetrodes (senior moment??), what you`ve said makes a lot of sense and indeed rings true with my findings.
as someone that learns best and finds most interest employing Empirical methodology I occasionally come up with "stuff" of interest (if only to myself) and am equally without explaination as to Why it exhibits X,Y or Z phenomenon, I Knew signing up to this place would pay dividends

I tried adding a Tuned front end and this worked equally well with the added advantage that when I moved to Earth plane wire on the RF meter towards the Antenna IN, the audio Shot up in volume and I was listening to Radio 4 LW as if it were a regular radio, the same when fed this into a triode (a CV455) and then amplified externaly.

it would seem that I accidentaly stumbled upon a crude Regen radio, talk about reinventing the Wheel

only difference of course is that this is a Single valve that will output to Speaker and do all the radio stuff at the same time.

Thanks for the inspiration Paul and Moose for the Answers I sought, this thread can be closed now.

27th Sep 2009, 5:06 pm	#1
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Pentode amp experiment. Inspired by Pauls post here: https://www.vintage-radio.net/forum/...81&postcount=2 stating that a Pentode could be used if wired/strapped as a triode, I tried this with a speaker on a home-made Valve Electronics set (like the old Science Fair sets but with Valve holders to the spring terminals) and decided to experiment a little. Firstly I took Pauls advice and wired the pentode as described, it worked but was pulling a LOT of power through even without an input, I use a small 250v 10W bulb in line with the HT supply as an indicator when I`m messing about so I tried all sorts of different configurations with these 3 grids to plays with, all connected togeter as one, closest to anode goes to anode and same with cathode, standard config and trying each as a Control grid etc... Many Many combos! and one I found that Does work and pulls very little, is g3 to anode, g2 to Cathode, and the g1 still as the control grid, although strapping g2 and g1 together makes little difference. I connected my Longwire to the control grid (g1) and checked for RF, sure enough it pinned the needle when it touched the anode connection! it seems to make a fairly reasonable RF amp as the longwire alone only moves the needle a little, all other configs do very little, the Above mentioned works great, and as I mentioned that g2 makes little difference straped to g1, for Audio this is the case but you get a little more RF taking g2 to cathode. what I`m acually hearing out of the speakers is the same as when you connect crystal set headphones to Ant and Gnd, only louder and NOT the same as connecting the same to the input of an audio amp no matter what the gain (I tried that too). I`m using an EL34, a 503 (0.05uF) antenna coupling cap to the control grid and pulling that down to deck with a 3.9M Ohm resistor. anode O/P is to the Mains in side of an 8VAC adaptor transformer from an old BT answering machine, feeding a 3 Ohm speaker. although I doubt any of that is Too Critical appart from using a high ohm resistor. does anyone know if this is a Normal/Known config for a pentode application? Woohoo! I finaly managed to figure out how to UL pictures too, here is my Valve Proto-typing set. Attached Thumbnails Last edited by YT2095UK; 27th Sep 2009 at 5:19 pm.

27th Sep 2009, 5:28 pm	#2
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,945	Re: Pentode amp experiment. Are you sure you've identified the pins properly? G2 (pin 4) should go to the anode and G3 (pin 1) to the cathode in a triode wired EL34. An EL34 is an odd valve to use in experiments like this since it is a very high power output valve. Something like an EF80 or, if you prefer octal valves, an EF36/37/37A/39 would work just as well and use much less power. I'm also unsure how an EL34 would perform at radio frequencies. It's also unclear how you're applying bias to the valve. Paul

27th Sep 2009, 5:51 pm	#3
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Pentode amp experiment. absolutely, pin for pin and double checked as always before I powered up. in fact your wiring is excatly the same as I used in the OP stage of my amp that I made a few weeks ago, only I used a 1K resistor to take g2 to Anode rather than directly wiring it. so I know what you said works as for Biasing, there is none, the Cathode is Hard to deck, no RC coupling at all. the Control grid (g1) is pulled down to deck via a 3.9M ohm, and fed with a 0.05uF cap. I`m using the EL34 because I know that worst comes to worst, it wouldn`t matter too much How I wired it, the power I`m using wouldn`t destroy it without me knowing about it in plenty of time before hand. as for RF on an EL34, I can say that LW is easily amplifiable (is that even a word), I`ll have to get the sig-gen out and try that, noting the drop off (I`m keeping plenty of notes!).

27th Sep 2009, 9:16 pm	#5
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Pentode amp experiment. it did take a lot, but the way I have it configured currently, it takes very little (the bulb doesn`t even show except a little in IR. g1 is the control as normal, g2 is direct to cathode, g3 is direct to anode. set up like This, I get a really good RF amplification (micro watts to milli watts) and that`s whilst it`s driving a speaker. now I realise that this in itself it`s practical for anything in particular as it stands, or with these particular parts I`v chosen to use, but I think there May be something useful here for Future work/experiments. incidentaly, I did try a coupling to the sig-gen and it`s ok up until about 3MHz and totally drops off at 3.5. I`ll replace the indutive load (OP transformer) with a resistor in the morning and try again, maybe the load plays a part?

28th Sep 2009, 12:05 pm	#6
Mr Moose Octode Join Date: Jul 2008 Location: Gloucestershire, UK. Posts: 1,043	Re: Pentode amp experiment. Hello, With g3 connected to the anode you are running the valve as a tetrode rather than a pentode, although with g2 connected to the cathode it will act as a suppressor grid as well as a screen grid. As Chris has said the anode current of an EL34 at zero bias and Va =250V, is over 250mA, however by connecting g2 to 0v you are probably reducing this to well below 25mA. By connecting g1 to the cathode with a 3.9meg resistor you have a grid leak detector, although 3.9 meg is a higher than the normal 1meg to 2meg for a grid leak, you might well be getting AF as well as any residual RF at the output. Yours, Richard

28th Sep 2009, 6:05 pm	#7
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Pentode amp experiment. I`m really glad you said that! I`d completely forgotten about Tetrodes (senior moment??), what you`ve said makes a lot of sense and indeed rings true with my findings. as someone that learns best and finds most interest employing Empirical methodology I occasionally come up with "stuff" of interest (if only to myself) and am equally without explaination as to Why it exhibits X,Y or Z phenomenon, I Knew signing up to this place would pay dividends I tried adding a Tuned front end and this worked equally well with the added advantage that when I moved to Earth plane wire on the RF meter towards the Antenna IN, the audio Shot up in volume and I was listening to Radio 4 LW as if it were a regular radio, the same when fed this into a triode (a CV455) and then amplified externaly. it would seem that I accidentaly stumbled upon a crude Regen radio, talk about reinventing the Wheel only difference of course is that this is a Single valve that will output to Speaker and do all the radio stuff at the same time. Thanks for the inspiration Paul and Moose for the Answers I sought, this thread can be closed now.