advertisement


My first foray into DIY valve electronics a.k.a adventures with a Velleman K-8020

James

James

Lord of the Erg\o/s
It's almost here, my new kit valve preamp. I just have to warm up the soldering iron and hope I don't make too much of a mess. However, seeing how this is a place where hardcore modders hang about, I was wondering about the wisdom of simplifying the design somewhat. I found this article where the circuit was changed from this ...

11757.gif


to this ...

11779.gif


Can anyone understand what has actually changed and if this is a sensible thing to do? I don't mind blowing up the kit preamp, but I don't really want to blow the rest of my kit up with it. Your collective counsel please.

James
 
There's an implication that the power supply voltage has also been changed. Might be worth looking for the detail on that before connecting everything up.

You've got a degenerated common cathode gain stage on the left and a cathode follower output buffer on the right. The mods delete the interstage capacitor and run the tube at higher currents. I would sit down with the ECC82 datasheet and google for circuit analyses. Here for example covers the gain stage. Then work out the gains, currents and voltages and see if it makes sense.

Paul
 
Paul Ranson said:
There's an implication that the power supply voltage has also been changed. Might be worth looking for the detail on that before connecting everything up.

You've got a degenerated common cathode gain stage on the left and a cathode follower output buffer on the right. The mods delete the interstage capacitor and run the tube at higher currents. I would sit down with the ECC82 datasheet and google for circuit analyses. Here for example covers the gain stage. Then work out the gains, currents and voltages and see if it makes sense.
Click to expand...
Paul,

Thanks for your comments. I'm afraid it's all a bit gobbly-gook to me at this stage. The author of the mods did say this ...

What is immediately clear is that the double triode ECC82 vacuum tube that is used is run rather conservatively with very little current flowing through it and sees a heavy load at the output in the form of a 5K6 resistor. This not only reduces its linear behavior but also causes for more distortion. Obviously this had to be changed; hence I started with calculating the proper current through both triodes and implementing the changes on the PCB. I also removed a number of components that served no real purpose, or only served to further complicate the schematic. The updated schematic has all the changes that I made and results in a compromise between lowest distortion and gain. It can be configured for even lower distortion, but the gain would rise considerably, which is not desirable.
Click to expand...

I take from his comments he knew what he was doing. All the datasheets on the ECC82 is not going to help me know if the modified circuit is stable, safe and not overly hard on the valves. My head is saying to go with the stock design and maybe swap in better ECC82s to see if I like how it sounds. My heart wants an ARC REF3 killer for a couple of hundred squids. I think I'll have to listen to my head.

James
 
That 5k6 does look odd. Is that actually what is supplied?

I'd leave the gain switch in place, it works by varying the negative feedback, it'll give you something interesting to listen for....

Paul
 
I think that 5k6 resistor is there as a dominant load so that the preamp output can drive a wider range of power amp loads without its performance varying much. The modded approach to that is the 1K series resistor on the output.

Anyway, I'd be inclined to build the thing exactly to the original spec, then try some tube rolling, and then look at some mod's. The modded schematic is quite a bit different to the original design.

Mr Tibbs
 
James,

The supplied ECC82s are probably Electro Harmonix (I don't have my pre infront of me to check); these are better than their reputation, reliable and relatively quiet. Philips JAN 5814s are a good low cost alternative (I'll try a pair later if you like, I've not bothered to date). Otherwise you're looking at NOS: Radio Technologie, Mullard etc. JJ, Telefunken and AEG ECC82/12AU7s all have high reputations but I've not tried any of them. My old Rose preamp came equipped with Golden Dragon 12AU7s and these seemed fine.

In general, the only stark improvement I've heard when tube rolling is when moving to Svetlana 'Winged C' EL34s. Gains must be very system and usage (and ear) dependent and, like most things in hi-fi, the real differences often over-stated.

PS - Just to echo Mr Tibbs: I would build the thing as the original designer intended and mod it (or not) from there.
 
Alex,

Good to hear from you, and thanks for your useful comments. I should add that the reason why I took your advice to try the Velleman was because I heard an Audio Research LS16 MK2 in my system when my B-250 was away for repairs. Whilst there was a lack of dynamic contrast and sheer energy compared with SS, there was a harmonic rightness and sense of cohesion that couldn't be denied. I hope the K8020 can give me some of that.

James
 
James, I'll be interested what you think - but don't forget its about a tenth the price of the Densen. It doesn't seem short on energy or dynamic contrast, although its a bit light-weight. The Border Patrol, which is in the same price range as the B-250, blows away any SS preamp I've heard.

Could you PM me how to get the valves out!? I didn't have the right allen key anyway, but it looks like I can't get the valves out without removing the sleeve, since the covers are in the way, and can't remove the sleeve without getting the valves out...
 
Alex S said:
James, I'll be interested what you think - but don't forget its about a tenth the price of the Densen. It doesn't seem short on energy or dynamic contrast, although its a bit light-weight. The Border Patrol, which is in the same price range as the B-250, blows away any SS preamp I've heard.
Click to expand...
Will do. I have to take delivery and build the thing first. I enquired about BP directly with the manufacturer. But I didn't get any response.

Could you PM me how to get the valves out!? I didn't have the right allen key anyway, but it looks like I can't get the valves out without removing the sleeve, since the covers are in the way, and can't remove the sleeve without getting the valves out...
Click to expand...
Oh dear. A pair of pliers and lashings of electrical tape for grip and protection against broken glass perhaps? I'll let you know when I have some idea what you're dealing with.

James
 
I've built the K8020, but unfortunately the right channel does not have any output. Both valves work fine (swapped them around to be sure), the power amps work fine (swapped them around to be sure) and I've examined my handiwork under a magnifying glass to check that I hadn't shorted any solder joints, broken any tracks, stuffed up polarity etc etc. Given that it powers up nicely and one channel works just fine, what on earth could have gone wrong?

Velleman suggested I use a multimeter and 'follow the signal' between the working and dead channel, but there are lethal voltages in there FFS. I'm brave but not stupidly brave. If it is a faulty component, what is it likely to be? Coupling cap? Relay? How can I tell?

Symptoms are: the right channel has zero output. Dead as a Dodo. Nothing to be heard from the right loudspeaker.

Mr Tibbs!!!

James
 
James, you need to run over the circuit with your multimeter while it is powered up and check the voltages with the diagram and between channels. Take care, but this really isn't dangerous, attach one end of your meter to 0v and just use one hand.

If the voltages check out then it pretty much has to be a coupling capacitor or the volume control.

Paul
 
Paul,

You make it sound so simple. I guess I never got over my brush with 240VAC in my youth. But I'm determined to give it a go. But first, a couple of questions.

I understand the part about attaching one probe to ground and using just the other to determine voltages. But do I need a steady input signal or will an idling circuit show up comparative faults? Based on the diagram on my initial post, am I looking for the voltages shown for the various points on the circuit? Would that be VAC or VDC?

James
 
I would first measure VDC with no input signal, and see if both channels have similar voltages. If not, you should be able to narrow the problem to one or two components, which you can check/replace/swap as appropriate.

If the channels are similar at DC, use a steady test signal (a signal generator is ideal, failing that a test CD, or any CD with fairly constant levels - thrash metal?) and measure the AC, looking for the point where it disappears.
 
Hi James

Follow the advice already given re checking DC voltages against the good channel. If that fails to bring any joy then you could try playing some music into the good channel, then take the output of the known working channel and 'inject' it into the faulty channel at various points and see if you can get some output.

Two obvious points to try injecting would be V1 terminals 2 and 7.

Use an old power amp and some dud speakers for the above test (say a NAP250 and SBL's arf arf)

Mr Tibbs
 
Mr Tibbs said:
Use an old power amp and some dud speakers for the above test (say a NAP250 and SBL's arf arf)
Click to expand...
Hello Mr Tibbs,

You should know better that I don't keep dud loudspeakers.

Anyhow, I've checked all the DC voltages and can confirm that they are present and correct in values according to Velleman spec. So it's not a power supply problem. Next, I traced the input to the ALPS volume control, which has 6 pins. The trio near the front is the left channel, and the trio near the back is the right channel. Looking directly at the messy side of the PCB, the signal goes into the left pins, exits via the middle pins and the right pins is fixed to 0V. So I measured the the resistance between input and output for both channels, and got identical readings. I then compared the output pins to the 0V pins and noticed that the right output pins was shorting to 0V.

A-ha! I have narrowed the problem down. I took the ALPS pot off the make sure it wasn't dud, and found that the track taking the attenuated signal to ECC82 pin #2 was completely and visibly isolated. I couldn't find where it could possibly short to ground. In desperation to fix the problem, I cut the suspect track (and confirmed it eliminated the short) near the ALPS pot, and hardwired the connection between the ALPS output and ECC82 pic #2.

I was 105% confident that that would fix the problem, but can you imagine my disappointment when the right channel remained dead as a Dodo? I shall have to take the PCB out again for the umpteenth time tomorrow and trace further downstream to see if there are any other shorts. So far, it hasn't been a very encouraging start into DIY electronics, but I'm learning plenty fast ...

James
 
You should know better that I don't keep dud loudspeakers.
Click to expand...

nuff said, James. Apologies for the flippancy during your hour of need ;)

A-ha! I have narrowed the problem down. I took the ALPS pot off the make sure it wasn't dud, and found that the track taking the attenuated signal to ECC82 pin #2 was completely and visibly isolated. I couldn't find where it could possibly short to ground. In desperation to fix the problem, I cut the suspect track (and confirmed it eliminated the short) near the ALPS pot, and hardwired the connection between the ALPS output and ECC82 pic #2.
Click to expand...

Be careful about attempting to measure resistance and continuity 'in circuit', as the various components (even capacitors) can give you a false reading (esp if using a digital meter). Try putting a music signal direct onto pin 2 of the valve (to bypass the input switch and volume pot) and see if you get an output.

Mr Tibbs
 
You must log in or register to reply here.


advertisement


Back
Top