Covid GM70 build

A Meanwell LRS-150-24 with the adjustment pot turned all the way down will heat GM-70s just fine, but you'll need the filament very close to AC/DC ground, so you have to ditch the cathode resistor and bypass cap and implement a fixed bias system instead to supply negative voltage to the GM70 grid.

I mention this because the Meanwell supplies are pretty inexpensive, not very heavy, and they are sufficiently quiet for use on a filament. I don't recommend getting any of the unshielded ones though!
 
The Meanwell looks good and I wish I knew about it much earlier. I bought and populated Rod Coleman's heater boards, his regulator boards, and the AnTek tranny all for the GM70 heaters. The Meanwell would have been much easier.

Is the a sound quality difference between a fixed cathode bias?
 
There has definitely been a preference for cathode bias over the years, though I wonder how much of that has to do with the limit of grid leak resistance when using fixed bias, and perhaps that being a little tougher for some driver circuits to use.

If you already have that stuff built, then don't derail your build! (by that I mean don't let *me* derail your build)
 
I use the Meanwell supplies for the GM70 but they aren't as good as a VCCS fed GM70. I used them straight into the GM70 filament to start with ... well via dropping resistors and CMC, supplies wound right down. I then added a VCCS regulator, different to the Coleman regs but similar idea and found that much better. I only went this route due to costs but also weight and size as mine is PP GM70. I think what you're doing is a good way of heating the GM70.

Fixed bias on these kind of valves saves a bit of heat but to get it right, is not as simple as it looks for all kinds of reasons; some as Paul highlighted.
 
I got an email from the post office saying my issue was "resolved" and I responded that I'd not recieved the package. They called me yesterday and said they could not find the package and "resolved" meant they did all they could with no luck.

I called Antek and they responded immediately by sending another transformer 1dt class priority mail.
 
Hi Paul. I mounted the toroidal GM70 heater tranny today and put 110v to it then measure mAC volts on the OT outputs.

The OTs have 4, 8, 16 outputs and the tranny has 2 sets of windings. So measuring all 12 combinations there was between 0.07 and 0.05 mAC coming out of the OTs

So, is the noticable hum? Or should I be ordering mu metal?
 
I just realized that I can hook the OT to one of the test speakers I have laying about. If there is hum, I'll hear it. It is one big toroidal.

20200610_201343.jpg
 
Ok, I need some help. When Kegger and I worked out this schematic he marked a couple resistors with watts and said the rest were .5 w or 1w or something like that. What watt should they be? P s, I added the 1W for the ones I dido know about.

location value value

Resistors

2 R2 220 1w
2 R3 1k 1w
2 R4 800 15w
2 R5 100k 1w
2 R6 22 1w
2 R7 22 1w
2 R8 510 1w
2 R9 510 1w
2 R10 68k 10w

Caps

2 C1 0.47 1000v
2 C2 50 200v
2 C3 220 10v
2 C4 47 ???


Kegger final version.jpg
 
R2 can be 1/8W.
R3 can be 1/4W.
R5 can be 1W.
R6/R7 can be 15W. I would change those to 47 ohm 10W.
R8/R9 can be 1/2W


I didn't notice how big R10 is. You don't really need the plate choke when you could already have such a giant dropping resistor, but having the plate choke would allow you to put in some over voltage protection for the 12GN7. When you turn the amp on, you will have 1300V appearing at the 9 pin socket, which is a recipe for disaster. You could put a string of Zener diodes in from the junction of the plate choke and the 68K resistor to ground to pull that down to 500V DC until the 12GN7 conducts and pulls it down further. This would give you a lot of safety benefit and would also protect your coupling cap, which will fail in this design if it's only rated for 1kV.

Is C4 the cap after the 68K resistor? Without Zener protection, this should be rated for 1.5kV. With Zener protection you could use a 500V rated cap (use Zeners to pull that node down below 500V at startup).
 
Thanks much Paul. The PTX is 1300v CT. After it goes through CLCLC it should be around 900v. I know in a few minutes and will post results.
 
Hey Guys, the amp fired right up without issues!!! Just power supply mind you but in this amp, that is the scary part. The PSU came up with 873vdc and she held steady at 907vdc. I assumed I would get lower v so I'm pleased with 907vdc.

I'm done for the night but in the morning I'll get the scope on her and see where the ripple is.

20200611_172951.jpg
 
R2 can be 1/8W.
R3 can be 1/4W.
R5 can be 1W.
R6/R7 can be 15W. I would change those to 47 ohm 10W.
R8/R9 can be 1/2W


I didn't notice how big R10 is. You don't really need the plate choke when you could already have such a giant dropping resistor, but having the plate choke would allow you to put in some over voltage protection for the 12GN7. When you turn the amp on, you will have 1300V appearing at the 9 pin socket, which is a recipe for disaster. You could put a string of Zener diodes in from the junction of the plate choke and the 68K resistor to ground to pull that down to 500V DC until the 12GN7 conducts and pulls it down further. This would give you a lot of safety benefit and would also protect your coupling cap, which will fail in this design if it's only rated for 1kV.

Is C4 the cap after the 68K resistor? Without Zener protection, this should be rated for 1.5kV. With Zener protection you could use a 500V rated cap (use Zeners to pull that node down below 500V at startup).

This is great Paul. But now I'm a bit confused. The B+ is 907vdc to the OPT and the GM70. my plan is to pull a line off the B+ and go to the 68k -10W resistors and cap to pull it down to 350vdc for B++ to the 9 pin.

So the 9 pin should only see 350 DC right? Or am I missing something?

PS, I really appreciate all your help Paul, it is invaluable!.
 
The cap doesn't pull anything down. A cap is 1,000,000,000,000,000,000... ohms at DC. The 12GN7 has to draw current to cause voltage drop across the 68K resistor. If you fire the amp up without the 12GN7, or when the 12GN7 is warming up, then there is no voltage drop across the 68K resistor, other than to charge up the cap, which won't take all that long.
 
I've been reading about star ground and how good it is keeping hum down however there seems to be a number of layouts. What is a good proven layout?
 
Back
Top