PP3521 idht

[simon]

My curve tracer broken no replies to emails. Is yours operational?test a few for me and buy some that suit you for you. I might buy a new kit but not right now.

Sadly I don't have one. One day maybe.

[Paul Barker]

Well theyre not going anywhere. Ill get one in time.

[simon]

[Mike H]

I should keep out of this, but...

Not sure anyone would argue that a unbipassed cathode resistor will produce lower distortion for a given output voltage. It is, after all, adding local negative feedback. The question should be does the higher anode resistance (with its attached lower PSRR) and lower gain (there is feedback after all) cause more or less problems that the bipassed cathode.

Where I will question the rationality however is if this is then used to justify the following chain of thinking.

1. Removing the bipass cap makes it sound better
2. The cap is the reason it sounds worst.
3. Capacitors make things sound worst.

If 1 is true, you have yet to prove 2, therefor we cant say 3 is true.

Got a BBC manual somewhere - how to design valve circuits - unbypassed cathode resistor increases anode resistance

[Paul Barker]

I know, its all factored in. Mu x rk added to internal resistance. Included in the gain calculation and the reason for the shift to the ecc88 because the gain is still sufficient and the internal resistance still low enough for a 10k load. In eror Id missed that the 10k load for the d3a wasnt sufficient. The ecc88 isnt a favourite valve in my life, I try to avoid it in favour of 5687, but the maths for the 5687 just not right. One could alter the whole design or use the ECC88. I have gone for the latter, because its worth adapting and seeing if I like the result.

In the event there is a delay finding out because the RS components order card declined Ive just had an email from RS a bit late in the day. Waiting for a time in daylight hours I can phone RS. As their email is offering no alternative to phoning. Seems AI isnt embedded in RS yet.

A bit of a downer becasue the values I ordered arent in ebay sellers lists of values for sufficient power rating. I have found ordering resistors on ebay is a most clumbsy experience, if you dont want the resistors they sell. Sounds obvious but gheir ranges are limited beyond belief. You see lovely pictures of beautifully presented chassis mount 25 Watt resistors as if taken by a professional photographer. But they may aswell be a photo of a cow pat with such limited resistance values listed. Id rather walk in a field avoiding cowpats than look for these values on ebay ever again.

[Paul Barker]

[Paul Barker]

After a phone conversation with RS, what a breath of fresh air talking to a human is if I might say so. Paid and gone through. My Nat west account is under scrutiny theyve asked for pin number far too often. RS order seen by fraud avoidance department since I havent ordered for years, as a danger. We also have had trouble getting Diana who has lost her mind internet banking again, which on android didnt work and I gave up, not worth the bother. Had a call from the fraud team asking if I had been trying to get internet banking ans a samsung blah blah, which yes I had. So I know they are scrutinising the account. I think theyre a bit too strong at trying to identify fraudsters. But I guess this is the future. Probably ai or morons identifying which transactions are fraud. But I vahuely recall a recent consumer program saying in about a year a government edict will come into play to ensure banks carry on refunding fraud, as now its a voluntary agreement only and there is so much fraud the banks are trying to swerve volunteering. In the interim, dont open and links sent bay any means.

[Paul Barker]

Made good progress today after rs order delivered. Mounted the power resistors and the two additional caps. Mounted OPT’s. All to do now is hook it up point to point. Put the b+ supply on a variac. Bring it up slowly test voltages if all good, plug in a source and audition. Maybe tomorrow end of day or next day, its not a competiton. Happy this far.







bp locations near me

[Mike H]

Some while ago I switched to Barclycard for RS, because normal bank card kept failing. Can't remember the exact reason, but probably same as you.

[Paul Barker]

Yes Ive had it before. Fraud departments are suspicious if RS.

[Paul Barker]

Well first turn on with variac. Only hearing one channel and had to turn varac to 150v dc as the two vr tubes are not engaging so the ecc88 sees the b+. So I have work to do to fault find. But I like the sound. So progress made and the amp when faults rectified will sound right at home.

Work to do though.

[Paul Barker]

A little confused whats going on, may be one for the output transformers causing all the problems. So I’ll have to unmount both and mount the 3k rCore into 8 ohm transformers which I know to be good.

Basically one channel is caining the output valves mainly known from the 360 ohm cathode resistor drop and the red plates on that side, and different voltage readings on opt that side.

Shucks, more work to do, but what is clear is the sound is sweet.

I suppose I only need to install one r core on the bad channel where the pp 3521’s are pulling excess current. If that fixes everything I can mount the other. Then I’ll have another pair of transformers to sort out on the chinese winder.

I wont become bored.

However it has just occurred one of the Hammonds might have received damage on the other amp which popped an opt dead short.

Oh well. Fortunately I have spare pair hammonds.

I discovered the ecc88 is not being mistret, so the vr tubes are striking. The reason the amp sounds sweet at 150v b+ is because the current drawer on the dodgy side doesnt go too excessive on the output stage.

[Nick]

Unless I am missing something, running without the output transformer and with the output valve to b+ should be the same as with the TX in place (for DC at least) to establish the op points.

What voltage is on the grid of the side that is having a problem? I would have expected that from a leaky coupling cap.

[Paul Barker]

Oh shucks it probably isn’t the opt, or the cathode resistor or any shorts. Nothing shorts dead testing including the bypass caps and the transformer and the cathode resistor. All resistance readings dead tested are perfect. Atransformer ohms primary and secondary test good. Neither wimding shorts to ground either.

That pretty well exposes one culprit, a output valve. Thats pretty poor of me not to consider that when I took them all out of the bases and put them all in the same bag, with no valve tester to hand.

I’ll have to make a test circuit to try each valve of these four in and see if its just one valve that draws the excesive current. As a pair are in parallel if there is an internal short of a grid pulling current. Maybe I could dead test the grid pins of each valve to each of the other pins ?

Im rather glad it doesnt seem to be the opt because im rather liking the sound. I recall doing a real effort winding them to my best ability, plenty of i ductance and plenty of inter layers, maybe eight you cant tell afterwards except by testing leakage inductance. But I think the bandwidth speaks for itself they arent limited.

Ill go back upstairs and see of ohms reading valve pins to each other reveals anything.

Another thought is though on dead testing all the capacitors seem seperate from ground doesnt prove they arent leaking to ground with voltage applied. Simple test for that would be take that channel’s bypass caps out of circuit because if its one of those that will fix it only leaving no base on that channel. I guess that will be the first live test. There might well be one duff cap which only shows its hand on a live test. I suppose with a test jig I could test them one at a time, starting with all disconnected. A time for clip leeds.

Why do I feel like Simon?

[Nick]

Well, if one channel works, you have at least two valves that work (or at least don't fail with high current), try running the two channels with a single valve each from the good pair. That will show if its the valves or something else.