A few years ago I bought a budget VR70e from a UK Ebayer. Apparently derated from 40 to 20watts to justify the low price (or poor quality EL34s?)
One EL34 has failed, but I am confused by the test point voltages. (test points are connected to the cathode)
The working channel's are both around 35mv but according to some other websites, should be 10 times this. The cathode resistors are 10R so the current is 3.5ma. The anode voltage is 450, so this means the valve is only dissipating less than 2 watts.
Or am I missing something?
I am intending to replace the 6F2s with ECF82s and fit a matched set of 4 EL34s as well as replace the output stage coupling caps with better ones. (The input caps I already changed to some Audio Note ones I had left from another project)
Ian
One EL34 has failed, but I am confused by the test point voltages. (test points are connected to the cathode)
The working channel's are both around 35mv but according to some other websites, should be 10 times this. The cathode resistors are 10R so the current is 3.5ma. The anode voltage is 450, so this means the valve is only dissipating less than 2 watts.
Or am I missing something?
I am intending to replace the 6F2s with ECF82s and fit a matched set of 4 EL34s as well as replace the output stage coupling caps with better ones. (The input caps I already changed to some Audio Note ones I had left from another project)
Ian
Yes I used my supercomputer to work out that the voltage across the cathode resistor should be 350mv, that's why I wondered why it was only 35mv!
It cost the same for a matched set of four, as individual non matched ones and I believe there is more to it than just being able to adjust the bias for equal current, such as gain, MU, transconductance GM & plate resistance Rp.
This formula seems to connect the 3 parameters than should be matched for triode connection:
[FONT=arial, helvetica]Mu = Gm x Rp[/FONT]
I am happy to put right on this, being at the bottom of the learning curve.
It cost the same for a matched set of four, as individual non matched ones and I believe there is more to it than just being able to adjust the bias for equal current, such as gain, MU, transconductance GM & plate resistance Rp.
This formula seems to connect the 3 parameters than should be matched for triode connection:
[FONT=arial, helvetica]Mu = Gm x Rp[/FONT]
I am happy to put right on this, being at the bottom of the learning curve.
3.5mA would suggest the tube is shot or something is wrong with its bias voltage. You are right about the matching parameters, but how many sellers would actually do it (without some decent markup), and how would you verify it without the proper test gear at home... And of course, to use a matched set, you should have OPTs with perfectly matched primary windings, what is the likelyhood that such transformers are included in an amplifier at this price point? 
I agree with your points and of course it means taking the seller on trust and that's something you cannot take for granted from internet businesses, sadly.
Also, surely the standing current mainly determines at what point the output changes from pure class A to AB? For sensitive speakers and / or low volume the amp should sound better if the standing current is high enough to keep it running in class A.
But if the speakers are not sensitive or always run quite loud, then it will be running in AB so could the standing current can be lower without affecting the sound?
Also, surely the standing current mainly determines at what point the output changes from pure class A to AB? For sensitive speakers and / or low volume the amp should sound better if the standing current is high enough to keep it running in class A.
But if the speakers are not sensitive or always run quite loud, then it will be running in AB so could the standing current can be lower without affecting the sound?
Re biasing this amp
I think I have the exact same amp supposedly derated to 20W per channel, sold on to me a few months ago through another eBayer. Considering how cheap it was, it sounds very nice. Having read batteryman’s experience from three years ago, I recently checked the bias voltages and current. I’m pretty sure it has some recently renewed EL34s so a check was due. To be (down) rated to 20W, I expected the settings to be set around 24mA. Before adjustment, they varied (valve to valve) from 39mA to 51mA so after adjustment, I set them all around 40mA (~520mV). Took some patience as adjustment took a while to settle down. I haven’t checked, but from schematics of the Dynavox version, I’m assuming the valve HT is around 420v, so that gives me around 34W per channel, well within EL34 tolerance and more than adequate for my listening. It’s a very decent amp, just missing its cage. I’d be grateful If anyone knows where I might get one. The guy who runs the company that sold this unit to its first purchaser is unlikely to sell me one! I’ll be taking it apart in the summer to mod for switchable triode connection. Thanks
I think I have the exact same amp supposedly derated to 20W per channel, sold on to me a few months ago through another eBayer. Considering how cheap it was, it sounds very nice. Having read batteryman’s experience from three years ago, I recently checked the bias voltages and current. I’m pretty sure it has some recently renewed EL34s so a check was due. To be (down) rated to 20W, I expected the settings to be set around 24mA. Before adjustment, they varied (valve to valve) from 39mA to 51mA so after adjustment, I set them all around 40mA (~520mV). Took some patience as adjustment took a while to settle down. I haven’t checked, but from schematics of the Dynavox version, I’m assuming the valve HT is around 420v, so that gives me around 34W per channel, well within EL34 tolerance and more than adequate for my listening. It’s a very decent amp, just missing its cage. I’d be grateful If anyone knows where I might get one. The guy who runs the company that sold this unit to its first purchaser is unlikely to sell me one! I’ll be taking it apart in the summer to mod for switchable triode connection. Thanks
Schematics please !
As regards to the low bias measured ( 35mV) ; measuring the cathode resistors will
answer what bias voltage ought to be. The cathode is connected to pin 8 in the socket,
the cathode resistor is connected between pin 8 and ground.
To measure, amp disconnected and cooled down, lift the EL34, connect the DVM between
pin 8 and chassies ( rivet that fastens the socket will do) and read the resistance.
If 1 ohm is then 35mV is reasonable as bias , 10ohm then 350mV is ok
As regards to the low bias measured ( 35mV) ; measuring the cathode resistors will
answer what bias voltage ought to be. The cathode is connected to pin 8 in the socket,
the cathode resistor is connected between pin 8 and ground.
To measure, amp disconnected and cooled down, lift the EL34, connect the DVM between
pin 8 and chassies ( rivet that fastens the socket will do) and read the resistance.
If 1 ohm is then 35mV is reasonable as bias , 10ohm then 350mV is ok
Yes, this error is shown in all published schematics.
I have used and given away one of these, but i never looked inside, and had
no schematics at the time.
I guess that the feedback will compensate for the unbalance but yes it should be
corrected.
Funny that the anode resistors are located on the powersupply board !
No idea why it was derated even if it has actually been done. I would have thought the easiest way to ‘derate’ would be to drop the bias current but that seemed relatively normal to me for an EL34 with ~450v on the anode plates although the measured 520 mV is a bit high for 38 mA current.
Thanks for the tip! I’ll try that but having set it to 38mA, which with my amp gives ~450mV suggests the cathode R is ~10 ohms. Should running in Class A for much of the time.
Any producer that tries to cut cost could do this. I'm talking about new production of an amp, supplying with cheaper transformers does make sens
if a reduction of performance is accepted.
But we have no real info if this reduction is actually done.
Measuring power output at 1000hz would give a clue, my rememberence of vr70
is that they produced 35w/channel just liek one would expect.
If power measurment is not possible, measuring the B+ ( or the plate
voltage on one EL34 at idle ) could be compared with the schematics. Rewinding the mains transformer to supply lower voltage is possible, but
the savings are only half of the saving cheaper output transformers will give.
The idle current has very little to do with the full power output in a push-pull amp, especially fix-bias.
I wonder if the "derating" means a higher load impedance for more linearity. Or just "rating" at a lower point on the rising THD curve. The same amp can be 1%THD at 40W or <0.4%THD at 20W. If the market favors low THD number rather than high Watt numbers, a maker could change the ad-sheet specification with no change in the amplifier.
When in doubt, it is rarely wrong to bias a decently-designed fix-bias pentode(/UL) amp around 70% of maximum plate dissipation. There is no "magic number" here, just a broad minimum which for "most" amps includes 70% Pdiss as a good point.
Taking the conservative 25W Pdiss number for EL34, we aim for 17W-18W. If plate voltage is 420, we want 40mA-43mA plate current. It is far safer to measure cathode current. This includes screen current. On EL34 at typical idle, Ig2 is about 15% of Ip. So we get a number like 48mA cathode current for "70% Pdiss".
I'd take 40mA or 52mA. I would not expect a huge difference. I'd like match to 20% or 10mA, but I have seen quite "poor" matches work very fine.
Using old-school 10 Ohm cathode resistor, this is indeed 400mV-520mV.
Actually, I assumed a plate voltage of 450 based on previous owner’s notes and I based my calculations for bias current on the 70% rule of thumb before following Weber’s guide. I was happy to note that the bias current I’d calculated was similar to that noted for a 450V plate voltage running at 70%.
Well, I'm the one who made the schematic in posting #10 and it hasn't got an error. The resistors in the split-load phase-splitter are NOT equal in the actual amplifier. Maybe they SHOULD be, but they're not.
The schematic came about this way:
I saw the amplifier VR 70e-2 advertized in Germany for an attractive price. I didn't want to invest in something I didn't knew how was built so I asked the shop for the schematic diagram. I didn't have high hopes for ever getting a response but in a couple of days the schematic was in my e-mail.
I have included here what I received and you will easily recognice the typical and flawed Chinese-style diagram with sloppy errors. I was disappointed but still I ordered the amplifier.
The amplifier arrived and I immediately took it apart. I traced all components and made the schematic diagram you've already seen. The anode-resistor of the P/I isn't 'in the power-suppy section' as suggested. From the original schematic you'll see it has just been drawn like it belongs in another section of the amplifier. It doesn't, as there is only one PCB carrying all components except for a small add-on board with extra powersupply filtering. This extra board btw, is what mainly sets the VR-70e-2 apart from the original VR-70e model.
The amplifier has seen much use since it was bought in 2011 and still carries the original tube-selection. Actually today I compaired it with a NAD-214 on my set with B&W DM610-S3 speakers. The NAD had a harsh, typical transistor-related aggressive sound. The VR-70e-2 has nothing of the sort. It's an amplifier you may listen to for hours and hours without tiring of the sound.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.