A couple months ago I posted an earlier version of this simple but fantastic 6V6 amp with (25ma constant current source) cathode-bias. but it was using a single supply for the op-275 opAmp.
I'm now using a dual +/- 20v supply.
Also, in this new version, you'll notice that the op-275 has higher input impedance by using the non-inverting input and now I was also able to eliminate coupling capacitors to the tube grids.
There are no grid leak resistors, of course, now either.
I encourage any of you to have fun building it.
The audio experience now seems even better than before. I believe it has to do with the direct coupling.
http://pbpix.com/amp/6v6%20pp%20amp%20change%20hw.png
I'm now using a dual +/- 20v supply.
Also, in this new version, you'll notice that the op-275 has higher input impedance by using the non-inverting input and now I was also able to eliminate coupling capacitors to the tube grids.
There are no grid leak resistors, of course, now either.
I encourage any of you to have fun building it.
The audio experience now seems even better than before. I believe it has to do with the direct coupling.
http://pbpix.com/amp/6v6%20pp%20amp%20change%20hw.png
Attachments
Last edited:
In my SHAG project also posted slightly more than 2 months ago )) there is no even cathode bias, no any bias.
You HAVE to have the grid leak resistor as the opamp can get damaged with its output internally disconnected so the tubes go unleashed, bump from the experience. In fact, to evenly balance the opamp's outputs you have to have 'em both, to negative and positive rails.
You HAVE to have the grid leak resistor as the opamp can get damaged with its output internally disconnected so the tubes go unleashed, bump from the experience. In fact, to evenly balance the opamp's outputs you have to have 'em both, to negative and positive rails.
So you are saying that the cathodes have a constant 25mA drain, the anodes will not vary, so no output.
Non starter comes to mind.
Not sure on the heater supply or voltage doubler. Why not use a proprietary transformer and safer power supply.
Non starter comes to mind.
Not sure on the heater supply or voltage doubler. Why not use a proprietary transformer and safer power supply.
How much power do you get out of this ?
You could also make the fixed bias by using CVS instead of CCS at the cathode.
Then it is possible to drive the output tubes to positive grid voltages (A2).
I have done this kind of construction with 6N6P dual triode.
Max. output is some 7 W. Without A2-drive it would be some 3 W only.
Attachments
You'd be right about non-starter... but you'll notice the pp cathodes are coupled audio wise with back to back 1000uf caps. The CCS is also bypassed for when it enters "B". Otherwise it's in "A"
25ma bias puts it in class "AB"
The power supply is very safe. .. very low ripple. But one can use any supply if they have improvement ideas. I like the simplicity of it all.
The power supply for my 6N6P PP and photo of first prototype.
Attachments
Last edited:
I didn't think the op-amp output balance was a problem, but I relayed your comment back to my friend who was advising me on the opamp section. He replied:
"I have no idea what "with its output internally disconnected" means nor have I ever heard of a 'warning' about such as occurrence.
Output offset is primarily determined by input offset. In this circuit output offset is nominally 35 mV and mismatch between the channels is less than 1 mV.
I could have gone to extreme measures to trim and balance it further but out of a bias of around 19 V it simply doesn't matter.
Tube variation can be as much as +-20% and this variation is 0.18% (taking the 35 mV number) with a .005% imbalance.
Even 1% bias resistors would be 5 times more than the opamp offset (which doesn't count the +-20% tube variation) and on the order of 200 times the imbalance."
That comment was about the possibility of damaging the opamp by the HV from directly connected tube. There might be a plate-to-grid short in a tube => the opamp gets fried.
Ok, ok, this is very unlike in the finished and working unit, but sheat happens once in awhile. And then you are trying to replace that tube and do not get what went wrong.
After all, I do not get what's so much the advantage of not having a Rg. .. not having any bias network would be much more an accomplishment IMO.
Ok, ok, this is very unlike in the finished and working unit, but sheat happens once in awhile. And then you are trying to replace that tube and do not get what went wrong.
After all, I do not get what's so much the advantage of not having a Rg. .. not having any bias network would be much more an accomplishment IMO.
There are very few tubes that can be used without any bias. Quickly comes to my mind 811A.
What is the operating of those 5687's in your SHAG ?
Take a look at schematics (there is the link in my post). I did not say "without any bias", I said "not having bias networks".
M-a-g-i-c.
M-a-g-i-c.
I'm also using ultra-linear output taps as well which reduces power.
So it's only around 3 watts RMS.
Now less than 7 Watts is dissipated in one 6V6 and g1 voltage is so high (-20V) that the tube can not be driven full.
You could get more output power by increasing the idle current of 6V6.
Concerning the need for grid (leak) resistor. The only reason I have those resistors in my amplifier is that I can remove the op-amp
if I want to test the tube alone.
Last edited:
Thanks... raising the idle current is a thought. However it's really plenty of power for my living room for sure. Big sound throughout the house... so it really can do the job. And the audio is fantastic.. rich and full.. 3 dimensional sounding.. very realistic.
To be honest I was not aware of the exact wattage until I measured it today w/a dummy load and fed in steady state 1khz signal...
(With +/- 20v Vcc, the op-275 op-amp can get a max of 34vPP to the grids.)
I have 4ohm speakers so I had a 4 ohm resistive load on it and measured about 3.5v RMS max just before distortion on the scope..
V sq / R
3.5 x 3.5 = 12.25v
12.25 / 4ohms = 3.06w
So that's 3 watts. ..
But I was just told by my friend that it isn't a valid test because the by-pass caps charge up and increases the bias... so output power is held back. And I did measure and verify that he's right. The bias voltage went up a couple volts.
So puzzled at how to get a true dynamic wattage measurement?
To be honest I was not aware of the exact wattage until I measured it today w/a dummy load and fed in steady state 1khz signal...
(With +/- 20v Vcc, the op-275 op-amp can get a max of 34vPP to the grids.)
I have 4ohm speakers so I had a 4 ohm resistive load on it and measured about 3.5v RMS max just before distortion on the scope..
V sq / R
3.5 x 3.5 = 12.25v
12.25 / 4ohms = 3.06w
So that's 3 watts. ..
But I was just told by my friend that it isn't a valid test because the by-pass caps charge up and increases the bias... so output power is held back. And I did measure and verify that he's right. The bias voltage went up a couple volts.
So puzzled at how to get a true dynamic wattage measurement?
Last edited:
With 295 V +Ub it is possible to get some 6...7 W from a pair of cathode biased (to 20 V) 6V6. This require 40... Vpp drive voltage.
Now the limiting component is the op-amp, but by increasing the anode current, i.e. by reducing the cathode voltage, you can get easily few more watts.
Now the limiting component is the op-amp, but by increasing the anode current, i.e. by reducing the cathode voltage, you can get easily few more watts.
You're right. W/the op275 using +/-20v Vcc I can only get 34vPP
If I jack it up to +/- 23 I can approach 40vPP output. But the max Vcc rating is +/- 22v.
There is another interesting chip I might experiment with. The OPA604 (single amp/chip)
It has a max Vcc of +/-24v. So perhaps I can get a bit more drive out of that.
And this opa604 has other nice features like only all FETs in the signal path.
But I don't know if it's a futile effort since doubling power is only 3db which is not perceptible.
... and I understand it would require at least 10db to get a perceptible difference.
btw:
I see that there is supposed to be a dual opa2604 .... but I can't find it at any of the large authorized distributors like Mouser or Digikey etc.
I see the dual ones for sale on Ebay from China But afraid.
... I've personally discovered that the Asian route is not a reliably honest source for chips... as they fake everything and sell mostly counterfeit chips.
I discovered this personally, when I bought some op275 chips from a China seller and when I used them I found the current draw was double the specs from the legitimate AD datasheet.
When I contacted the seller with the measurements, he actually admitted they were fakes and returned my money advising me to trash the fake chips.
If I jack it up to +/- 23 I can approach 40vPP output. But the max Vcc rating is +/- 22v.
There is another interesting chip I might experiment with. The OPA604 (single amp/chip)
It has a max Vcc of +/-24v. So perhaps I can get a bit more drive out of that.
And this opa604 has other nice features like only all FETs in the signal path.
But I don't know if it's a futile effort since doubling power is only 3db which is not perceptible.
... and I understand it would require at least 10db to get a perceptible difference.
btw:
I see that there is supposed to be a dual opa2604 .... but I can't find it at any of the large authorized distributors like Mouser or Digikey etc.
I see the dual ones for sale on Ebay from China But afraid.
... I've personally discovered that the Asian route is not a reliably honest source for chips... as they fake everything and sell mostly counterfeit chips.
I discovered this personally, when I bought some op275 chips from a China seller and when I used them I found the current draw was double the specs from the legitimate AD datasheet.
When I contacted the seller with the measurements, he actually admitted they were fakes and returned my money advising me to trash the fake chips.
Last edited:
It drives 600R loads, for some setup it might help to get a bit into the grid-drive area.
Let us know if you find the reliable source.
I used NE5532 in mine, +/-22 Vcc max, also can drive 600R.
There is OPA445 with up to +/-45 Vcc.
Let us know if you find the reliable source.
I used NE5532 in mine, +/-22 Vcc max, also can drive 600R.
There is OPA445 with up to +/-45 Vcc.
- Status
- Not open for further replies.