Yes, it should be, but I was asking the OP what he measures because the tubes aren't biasing properly.
The "manufacturer-recommended" method was intended for hobbyists in 1960 who didn't have access to expensive test equipment. It also says nothing about checking the actual operating current of the tubes, which is important with modern 7591s. Today anyone can buy a digital multimeter for $20. There's no reason to drag a tuner and a loudspeaker onto the test bench, listening for "clicks" and "hum" and guessing what the actual bias is. And adding two extra test points can make the process easy and safe.
Thanks for all the replies. I'm a complete newbie to tube amps, so I appreciate hearing different opinions and approaches. It's actually helping me understand the amp a bit better.
When I got this amp it was already modified from the original configuration with one 3.3ohm resistor for each pair of 7591 tubes. I don't think it makes sense to revert back to the original design at this point. So I need to learn how to set Balance and Bias with the four cathode resistors (2ohm 5% now, but I may replace those with the 1ohm 1% I purchased if it makes sense). I've seen mentions of people adding test points on top of chassis to make things safer and more convenient, but I haven't been able to find a good explanation about how to do this yet. Definitely something I'd like to consider in the future.
When you say to check "bias voltage at the output tube grids", do you mean test voltage between Pin 6 on each 7591 tube and chassis ground? I will also look into checking voltages in the negative bias circuit, but not too sure how to go about that now. If you can provide more details I'd appreciate it.
I'm also going to try swapping the 7591 tubes from Right channel to Left to see if the lower bias voltage follows the tubes. I have another set of vintage tubes as well that should be in better condition, but I wanted to make sure the amp was working properly before I put them in.
Thanks again,
Andy
Last edited:
I just swapped the tubes from right channel to left and the bias followed the tubes. Last night bias on right was about 60mV max and left was around 40mV. Today that is reversed after moving tubes.
I also checked grid voltage between pin 6 and ground on each tube.
V4 -29.28VDC
V5 -29.00
V104 -26.50
V105 -27.13
Is that an indication that the tubes are the problem? The amp seemed to be working fine when I got it and it sounded good to me with those tubes. One does show some signs of possible overheating as seen in pic below. That tube is in channel with the lower maximum bias. I did swap that for another tube that looked better, but sound was a bit distorted and there was some hum. So I switched back and it sounded better.
Any feedback is much appreciated.
Cheers.
I also checked grid voltage between pin 6 and ground on each tube.
V4 -29.28VDC
V5 -29.00
V104 -26.50
V105 -27.13
Is that an indication that the tubes are the problem? The amp seemed to be working fine when I got it and it sounded good to me with those tubes. One does show some signs of possible overheating as seen in pic below. That tube is in channel with the lower maximum bias. I did swap that for another tube that looked better, but sound was a bit distorted and there was some hum. So I switched back and it sounded better.
Any feedback is much appreciated.
Cheers.
Are the two voltages circled below on either side of the 33W resistor what you're talking about? Would those be checked with positive lead of DVM clipped to resistor and negative to chassis?
Sounds like those 7591s are a bit tired, but it also appears that your bias voltage is too high. All those components in the bias supply/heater string need to be checked (amp off, of course). Then measure the voltages, red lead to voltage point and black lead to ground (you'll see a negative reading on the meter). The 12AX7 heaters should be 11 volts each, that's the way Scott ran them.
Has the selenium bridge rectifier been replaced? If not it should be, with a standard 100VDC+ bridge. The 18 ohm resistors are prone to drift, usually high but who knows.
In the pic below, note that someone has paralleled the circled resistor, one is an original carbon comp and the other is a new metal film. That's suspicious, and might be the cause of the elevated bias voltage. Also those copper-colored caps are bulging and should be replaced.
Has the selenium bridge rectifier been replaced? If not it should be, with a standard 100VDC+ bridge. The 18 ohm resistors are prone to drift, usually high but who knows.
In the pic below, note that someone has paralleled the circled resistor, one is an original carbon comp and the other is a new metal film. That's suspicious, and might be the cause of the elevated bias voltage. Also those copper-colored caps are bulging and should be replaced.
Wow, good eye! Thanks for catching that resistor. There are two 22K in parallel. It's definitely odd, so I'll remove them so I can test them and see what's going on. There's a better pic below.
I believe the rectifier is a more modern silicon version. Another pic below.
I also have new electrolytics to replace those copper ones, and the other larger one. So I'm probably jumping the gun by trying to set bias now, but it's a learning curve and I'm trying to do this in small steps with testing along the way so I can catch any mistakes.
I'm not sure what you mean about checking heater voltage on 12ax7s, but was doing some research. Is that pin 5 on the tube socket?
Thanks very much.
I believe the rectifier is a more modern silicon version. Another pic below.
I also have new electrolytics to replace those copper ones, and the other larger one. So I'm probably jumping the gun by trying to set bias now, but it's a learning curve and I'm trying to do this in small steps with testing along the way so I can catch any mistakes.
I'm not sure what you mean about checking heater voltage on 12ax7s, but was doing some research. Is that pin 5 on the tube socket?
Thanks very much.
Last edited:
Ok, so selenium rectifier replaced, good! Those caps look marginal, though. Don't remove the paralleled resistors, you'll lose the bias voltage altogether. Snip the metal film one temporarily and see how much bias voltage you get. My guess? Someone did that to increase the amount of bias voltage available for modern Russian 7591s. By reducing the amount of resistance leading to the bias pot, they increased the negative voltage available to the grids of the output tubes.
The 12AX7s should read about 11 volts DC across pins 4 and 5. I think that extra resistor was added to accomodate new-type 7591s. Snip the added one and you'll probably be fine. But I'd leave it in place in case you want to resolder it for new-made 7591s.
The 12AX7s should read about 11 volts DC across pins 4 and 5. I think that extra resistor was added to accomodate new-type 7591s. Snip the added one and you'll probably be fine. But I'd leave it in place in case you want to resolder it for new-made 7591s.
Last edited:
Great advice! Thanks again. I didn't mean I'd remove resistors permanently. Just to check the resistance and then replace, but snipping one lead makes a lot more sense. I see the 12AX7 tube numbers with the 4 & 5 pins on the schematic now too. I wasn't sure what that part was all about before now.
I clipped one lead on the 22K metal film resistor today. That one tested 21.94K and the older one under it tested 17.11K, but that was in circuit, so not sure if that's an accurate measurement.
After clipping the lead I rechecked the grid voltages on Pin 6 of the 7591 tubes and it changed as below:
V4 Before -29.28 After -25.60
V5 Before -29.00 After -25.25
V104 Before -26.50 After -23.10
V105 Before -27.13 After -23.60
Then I rechecked the bias (still with same tubes). Right channel increased from a max of 40mV to 60mV and Left increase from max of 60mV to 80mV.
The voltage at R209 33Ω resistor was -61.8V on right side (schematic says -56) and -56.5 on left side (schematic says -51.5).
Also checked those 18Ω resistors on C203. R206 measured 20Ω, but R207 and 208 both measured 18Ω.
Finally, I checked the heater voltages on the 12AX7 and they were all about 12V.
Any thoughts on those results?
Thanks again for any and all feedback.
Cheers.
After clipping the lead I rechecked the grid voltages on Pin 6 of the 7591 tubes and it changed as below:
V4 Before -29.28 After -25.60
V5 Before -29.00 After -25.25
V104 Before -26.50 After -23.10
V105 Before -27.13 After -23.60
Then I rechecked the bias (still with same tubes). Right channel increased from a max of 40mV to 60mV and Left increase from max of 60mV to 80mV.
The voltage at R209 33Ω resistor was -61.8V on right side (schematic says -56) and -56.5 on left side (schematic says -51.5).
Also checked those 18Ω resistors on C203. R206 measured 20Ω, but R207 and 208 both measured 18Ω.
Finally, I checked the heater voltages on the 12AX7 and they were all about 12V.
Any thoughts on those results?
Thanks again for any and all feedback.
Cheers.
You're definitely more in the ballpark. The replacement bridge probably has less voltage drop than the original selenium. You could probably increase R209 to 40 ohms to account for that. Or you might be fine where you are. Ideally you want around 22VDC on the output tube grids. But those 7591s are not well matched and you might have trouble getting them up to speed. You really ought to get a new set of Tung-Sols before you do any further modifications. Right now you're trying to accomodate those old tubes and that's less than ideal. :-(
That's good to hear I'm in the ballpark. I have another set of vintage 7591s that I will try soon. I didn't want risk damaging them by putting them in before I knew things were operating reasonably well.
Can't thank you enough for your assistance! I would have never been able to figure this out on my own.
Can't thank you enough for your assistance! I would have never been able to figure this out on my own.
It would be good to see if the other 7591s bias a bit more consistently. Given the way the voltages are now, I don't think there's any danger in giving them a try.
I replaced those copper coloured caps tonight and put the other tubes in. These are the grid voltages I got after that.
V4 -22.08
V5 -22.39
V104 -23.40
V105 -25.50
Any thoughts on why V105 would be higher? The tube seller marked them with test numbers and V105 was the highest at 9000. The others were 7000, 7500, and 8000, but I'm not sure what those numbers mean.
The max right bias was 85mV and max left was 65mV. That's an improvement over the other tubes and I was able to set all four to 60mV, which should equal 30mA with the 2ohm cathode resistors.
I'm listening to the amp now and it's sounding pretty good 😊
V4 -22.08
V5 -22.39
V104 -23.40
V105 -25.50
Any thoughts on why V105 would be higher? The tube seller marked them with test numbers and V105 was the highest at 9000. The others were 7000, 7500, and 8000, but I'm not sure what those numbers mean.
The max right bias was 85mV and max left was 65mV. That's an improvement over the other tubes and I was able to set all four to 60mV, which should equal 30mA with the 2ohm cathode resistors.
I'm listening to the amp now and it's sounding pretty good 😊
You're adjusting the balance pot, right? The key is to match the current draw of all four tubes as measured across the the 2 ohm cathode resistors. Depending on how well-matched the tubes are, you might get a different reading at each grid. That's a function of each tube requiring slightly different bias votage. It's current matching that you want for lowest noise and distortion.
Yes, I'm adjusting the balance pot. I started by setting balance to zero volts. Then set bias, then went back and forth tweaking balance and bias until all four tubes measured 60mV.
Thanks. I'll keep working on it and will continue replacing the old Ceracaps. I also have a CL-80 Inrush Current Limiter to install as well as a Safety Capacitor. Can't remember where I read it, but when I was putting together my parts list I saw a recommendation to replace the .01uF 10KV C206 capacitor with a .01uF 400/250V Safety Cap shown below. Curious if you have any thoughts on that.
Cheers.
https://www.mouser.ca/ProductDetail/80-C981U103MYVDBAP
Cheers.
https://www.mouser.ca/ProductDetail/80-C981U103MYVDBAP
If I'm not mistaken it looks like you have all the original aluminum electrolytic capacitors. I would replace them as they are now almost old enough to collect Social Security retirement benefits. There are differing opinions on whether or not you can reform them but I prefer to replace them.
You can find replacement multicap can filter capacitors at Antique Electronic Supply.
https://www.tubesandmore.com/products/capacitors?filters=2341a2356
Some people use Hayseed Hamfest products, they have a set that gives you what you need.
https://hayseedhamfest.com/products/h-h-scott-lk-72b-re-cap-kit-4-cans
I restuff the originals with individual caps inside. The way I do it makes them indistinguishable from the original. It's a bunch of work but a lot cheaper, especially if you're doing a LK 72 or a 299.
Keep in mind that the filter caps for the bias supply/heater circuit are common positive given that the bias supply is a negative voltage. I'm not sure if AES has those, Hayseed Hamfest does.
You can find replacement multicap can filter capacitors at Antique Electronic Supply.
https://www.tubesandmore.com/products/capacitors?filters=2341a2356
Some people use Hayseed Hamfest products, they have a set that gives you what you need.
https://hayseedhamfest.com/products/h-h-scott-lk-72b-re-cap-kit-4-cans
I restuff the originals with individual caps inside. The way I do it makes them indistinguishable from the original. It's a bunch of work but a lot cheaper, especially if you're doing a LK 72 or a 299.
Keep in mind that the filter caps for the bias supply/heater circuit are common positive given that the bias supply is a negative voltage. I'm not sure if AES has those, Hayseed Hamfest does.