JTM45 Build with various output tubes and GZ34/Solid State Rectifier

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
yep as i thought, had to be connected, now getting 465v on B+ wth diodes, will leave on now and see how hot the mains txf will get.
started at 23C its now gone up to 27C after about 5 minutes.
first 32uf cap connection is at 38C?
could it be a dodgy 32uf cap?
 
Last edited:
ive used a bias calculator which was provided in this thread, and it said that for EL34 at 70% it should be 43mA with 403v which is what i was getting.
is the bias calculator wrong?

Im now at 443v B+ so will adjust bias to 39.5mA

BTW. have ordered some KT66.
 
Last edited:
One of my concerns was it was getting rather hot, my only other concern is the bias, i am using a 68k for the 180k into the diode and a 5.6k then a 27k with a 50k pot, should i be concerned that this is vastly different than the regular setup of 180k into diode then 15k, then 56k into 50k trimmer?
 
Last edited:
so a matched pair of KT66 have arrived to day and i thought, right , i'll test the voltage on CT which was 433v and use the tube calculator for bias.
i could see that both EL34 and KT66 needed 40mA for 70% A/B.
so basically i had it on 39mA with the EL34s and would just need to increase 1mA once i'd put the KT66 in.
here is where it gets weird, so i the previous post i put that my bias is now 68k into diode into 5.6k into 27k into 50k trim pot and im getting 39mA using EL34s.
Ive now put in KT66 matched pair and im getting 170mA and can only lower to 70mA?
so i thought ok well i'll change the bias resistors for the more regular values used on the original schematic, but first i'll change the 68k to 150k, now bias at lowest point is 180mA so thats gone completely the wrong way?
i want to understand whats going on before i continue?
 
Do you understand a potential divider? In a potential divider, when you make the series resistor bigger, you are making the output voltage smaller. It's like turning a volume pot down; you get less voltage out of it (i.e. the output gets closer to 0V). With a smaller bias voltage going to the tube grids, they bias hotter.

I admire your ambition, but you seem to be out of your depth with the level of complexity you're attempting.
 
Last edited:
= V out in this case 41.4ma
Volts are not amps...

Step 1: Take the damn tubes out.
2: Get the bias supply working so the bias voltage is adjustable from, let's say -10V down to -45V or so. That's enouge range to cover any tube type.
3: Put the tubes back in and bias them.
4: Stop pratting about with the sodding bias supply from then onwards.
 
I have installed 150k into diode (1n4007) into 15k then 56k (voltage divider) then into 50k trimmer to gnd
at 150k / diode i get 115v
at diode / 15k i get -55v
at voltage divider i get 43.3v
when i measure the mA on the 1R resistors its at 180mA? not 40mA?
 
ive changed the 150k now to 56k and im getting -43v max on bias, i dont know whats happened to the 1R readings on pin1/8 to gnd?

Ah i see my problem i think, im tapping off between the 56k and the 50k trimmer as a voltage divider rather than from between the 15k and the 56k!!!
 
well its much closer now, ive put back the 150k into the diode and then have 15k and 56k with 50k trim down to gnd, im now tapping off the 15k/56k voltage divider!!! and getting -43.4v but when i measure mA on the 1R resistor on pin 1/8 to gnd its measure 68mv (this is meant to reflect the same as the bias but read as mA)
 
Text descriptions are hard to follow, so maybe I misunderstand you, but just in case:
You are aware the the bias current in the tube is not numerically the same as the bias voltage on the grid, right?
If you're aiming for 41mA in the tube (41mV measured across 1 ohm), do not expect the bias voltage to be -41V. The grid voltage and the bias current are two completely different numbers. You simply adjust the voltage to whatever it needs to be to achieve 41mA in the tube .
I don't mean to patronise, just better to clear this up.
 
thats exactly what im looking for, im confused of what readings to follow and you have just said that the mv reading on pin1/8 is what i need to read and not worry about the -43v on the bias.
ive lowered the 150k to 133k but i need to go a little lower so ill try a 100k in there and it should be about right hopefully!!!

then i can leave you all alone for a while!!!
 
Volts are not amps...

Step 1: Take the damn tubes out.
2: Get the bias supply working so the bias voltage is adjustable from, let's say -10V down to -45V or so. That's enouge range to cover any tube type.
3: Put the tubes back in and bias them.
4: Stop pratting about with the sodding bias supply from then onwards.

This is all you need to do and then forget about it forever

I have installed 150k into diode (1n4007) into 15k then 56k (voltage divider) then into 50k trimmer to gnd
56k is not on the schemo. You must mean the 47k, which is part of the second leg of the divider. It keeps you from shorting to ground through by setting the trimmer to 0v. 150k + 15k is your one leg of the divider, 56k + 50k trimmer is the other. Think of the 150k and 15k as one resistor, and the 50k + 56k as a second resistor.

ve lowered the 150k to 133k but i need to go a little lower so ill try a 100k in there and it should be about right hopefully!!!

It doesn't matter what value resistors you put in the bias supply. The goal is that without the tubes installed, you can adjust the trim pot to give you something like -10v when it's one direction, and -45v when it's all the way the other direction. It doesn't matter exactly, as long as you have a range to work with.

The idea is that every individual tube needs a different voltage on the grid to flow the correct current at idle, in this case 41ma. An EL34 might need -39v on the grid, but a KT66 might need -45v grid to get the 41mA at any given B+. If you crank the B+ up, you will automatically flow current through the tube, so you lower it back down with more (bigger number) negative voltage on the grid. That's why you set the bias range like merlin said and the forget about it except when you bias tubes. You only use the trimmer because different tubes only need a slight difference in voltage.


2: Get the bias supply working so the bias voltage is adjustable from, let's say -10V down to -45V or so. That's enouge range to cover any tube type.

This is the way to do that:

Think of the 150k and the 15k as adding up to one big 165k resistor since they're all both in series. Then there is the separate 50k + 56k to ground. The series resistor and the trimmer make up the two legs of your voltage divider; look up videos or articles on the web, however you learn the best.

Your voltage divider is the 165k resistor and the trimmer resistors. Diode doesn't matter for what you're doing here. Ignore the trimmer for a second and just focus on what the 165k is doing. It sets the maximum voltage you can put on the grids of your tubes. If you put 150v- into the 165k from your transformer, you might have something like 120v- coming out on the other leg. If you make this SERIES resistor BIGGER, it lets LESS voltage (LOWER NUMBER) through to the divider. A 400k instead of 165k might give you something like 90v- instead of 120v-.

-----------------

Your SERIES resistor set the MAXIMUM voltage (BIGGEST NUMBER) that can be applied to the grid. and your TRIMMER resistor LOWERS the voltage down to 0v- from the maximum.

So, what you do is shrink the 165k to a certain value so you get around 50v- on the "output" leg. This is your maximum negative voltage from the bias supply.

Then you TURN DOWN this voltage by making the trimmer resistance SMALLER. When your trimmer is all the way up you have 50k ohms. As you turn the trimmer DOWN you get a SMALLER resistance. This lets more of your maximum voltage go to ground. This way you have less voltage going to the grids of your tubes.

No tube in this amp is going to need more than -45v or -50v on the grids so that's a good number to shoot for.
 
Take your tubes out, set the trimmer to 50k.

Shrink the biggest series resistor first, then lower the value on the little one if you need to fine tune it. Put in a 100k 3w and 15k and see what your voltage is. If it's like -80v then you need to go lower on the 100k. If it's -55, put in a 10k instead of 15k. Not life or death here just get it close to 50. Within 10v or so.
 
thats exactly what im looking for, im confused of what readings to follow and you have just said that the mv reading on pin1/8 is what i need to read and not worry about the -43v on the bias.
ive lowered the 150k to 133k but i need to go a little lower so ill try a 100k in there and it should be about right hopefully!!!
The 43 volts is what gets you the correct milliamp rating at the 1ohm resistor. If you have 44ma at the resistor, you need to increase the -43v to more voltage (-45ish) to lower the current flowing through the tube to 41ma.

You're applying VOLTAGE at the GRID to limit CURRENT flowing through the CATHODE.

"About right" is the correct way to think of the big resistor in the bias supply. Set it like Merlin said and forget it. Tune with your trimmer, and don't worry about the 56k.
 

Latest posts

Back
Top