advertisement


A valve line pre amp for the DIY fishy

George J said:
Dear Arkless,

I have tried and failed to find a modern pure mono pre-amp with only mono inputs and output. Would you be prepared to make one such based on you design here published, but using valve rectification? It will drive either a Quad II Forty or Leak 12.1 onto a single ESL. So issues of the output capacitor are nothing to consider being careful with.

Sorry for the thread diversion.

Best wishes from George
Click to expand...

No problem. PM me with more details of your requirements if you wish.
 
Quick question - where can I get a bracket to mount the valve properly? I can bodge it in the short term, but cannot find anything specific for if I want to make it nice...my google foo seems weak!

Thanks, Richard
 
Take one 22mm hole saw, drills, files, hack saw and some aluminium.... or you could mount the valve(/s if using valve rectifier) on top and on view or on a shelf inside...
 
''where can I get a bracket to mount the valve properly?''
I made my own from a scrap of aluminium. Q-max hole punch (3/4'' or 19mm for most B9A bases, but check yours.) is neatest and easiest, like these https://www.cromwell.co.uk/shop/hand-tools/punches-wad-and-hole/hole-cutters-imperial/f/7111 or http://rover.ebay.com/rover/1/710-5...0001&campid=5338728743&icep_item=142477127321
Any way you do it make sure you cut the hole first then the folding and finishing!
I like Yanks suggestion, but then you have no scope to rotate the base to align the pins for the best layout. (Heater pins 4 and 5 closest to chassis for example.)
 
This site contains affiliate links for which pink fish media may be compensated.
On a general note, I have been listening to mine over a few days to get used to it before making any changes. (I started with 90k and I'm now on 180k for R4/10 as I do not need so much gain.)
I have tried it with my AI500 repro, Stereo 20 and old NAP110 and it performs well with them all in standard configuration. I rather like it. I do not notice any bass issues with 2.2uF output caps. The only thing I have played with is the valve. I have Brimar, Mullard and a 'Colmor' branded Mullard (with different structure) ECC88s. Each one has a different sound. The Brimar is nice, but tilts everything to the upper frequencies. The Mullards are smoother with the Colmor just a touch in front, but is also the worst one for being microphonic...
I have also tried a Russian 6N1P in there and that too is very nice and very cheap.
 
I couldn't find IN5953BG zeners on Mouse - I selected 1N5383BRLG instead - is OK? 150v 5W zener.
 
It's not ideal no. You can go down from the 1.5W rating but going up to 5W is not a good idea due to it's high junction capacitance. It's really non critical overall and any Zener diode of 120 - 160V will do but keep the power rating down. It is only conducting whilst the valve warms up.
 
snowman_al said:
I stand corrected... But how does the extra capacitance effect this application?
Click to expand...

Whether of not the effect would be pronounced enough to be audible to even the golden eared is debatable, but it is a non linear, voltage dependent capacitance and will cause some degree of HF even order distortion.
 
A random question, to understand the circuit better.

If C5 / C6 were much bigger, the HV would rise more slowly, reducing the turn-on thump, so that the constraints on the output coupling caps could be relaxed. Is this right?

Obviously, large value caps at 450V rating are big and expensive, so this isn't a practical solution, but just checking I have read the schematic correctly. At present they give a time constant of 15x47ms, about 0.7s. If they were 10 times bigger, the HV rise would be 10 times slower, and the thump reduced to a tenth the size.
 
PigletsDad said:
A random question, to understand the circuit better.

If C5 / C6 were much bigger, the HV would rise more slowly, reducing the turn-on thump, so that the constraints on the output coupling caps could be relaxed. Is this right?

Obviously, large value caps at 450V rating are big and expensive, so this isn't a practical solution, but just checking I have read the schematic correctly. At present they give a time constant of 15x47ms, about 0.7s. If they were 10 times bigger, the HV rise would be 10 times slower, and the thump reduced to a tenth the size.
Click to expand...

Indeed this could be done but the size of electrolytics needed to render the issue cured, especially if using 10uF + output caps, would get a bit silly and could have consequences elsewhere such as the dielectric absorption of such big caps potentially having a negative effect on SQ.

A valve rectifier is another much better and more practical option but would then ideally need twin heater windings on the mains TX (and a centre tapped 500V HT winding for bi-phase rectification) so the rectifier heater/cathode can be floated up at 350V and get around heater to cathode insulation issues. It is also not foolproof as if one switches it off, but then changes ones mind after 20 seconds and switches it back on, the still hot rectifier heater may allow the HT to rise too quickly... The triode will also still be warm of course which ameliorates this.

The complete solution is the SS delay circuitry which grounds the output for a minute to allow the valve to fully warm up and the output caps to fully discharge before opening the output relay. As I said earlier in the thread I'll publish this if there is a demand.
 
Thanks All - have now ordered some BZY97C (150v, 1.3W) - I am happy to run without these, but also wanted to add them and see if I could detect a change in SQ...now I can compare 5W and 1.3W versions :)

Oh, I have a couple of variable valve rectifier PCBs too, but want to start with basic cct first and take it from there.
 
Richard,

I'm making a bracket similar to the ones pictured for another forum member. Would you like me to knock up another and mail it over?
 
Cheers Mark - 19mm, 22.5mm and 25mm hole cutters should be with me by end of next week now, and I can get some suitable Alu sheet from a DIY store locally, so I’m quite interested to try myself (I just never knew how before :)).

If you still don’t mind doing one, I’d still say yes please...just in case I feck mine up :) I’ll cover costs of course - do you still have my address, or should I PM you?

Richard
 
Dowser said:
I couldn't find IN5953BG zeners on Mouse - I selected 1N5383BRLG instead - is OK? 150v 5W zener.
Click to expand...
Two 75V in series would be better. Shares the power, lower voltage zeners are closer to "ideal" and the capacitance is in series, so reduced
 
That had in fact occurred to me but I reckon if starting to get that worried about it then go for the relay delay and 400V output caps and be done with them altogether:).... the effect of the capacitance is likely to be inaudible, but a 5W rated one was pushing it! The last thing we are worried about here is ideal Zener performance, it just has to conduct at plenty above the 70V or so anode potential so as to be well out of circuit during normal operation but soon enough to allow 150V rated output caps and limit the degree of the switch on thump. 400V+ rated polypropylene caps are readily available if people want to get rid of the Zeners altogether but I would not do so if not using the output mute circuitry or a valve rectifier.
 
I intend to try without the zeners first anyhow, measure the output with a scope, and try first by powering amp (currently ML No. 29) on afterwards. The only problem I’ve ever had was with an XTC POW-1 power amp I have, it was fine with its matching PRE-1 (with 20 second power on delay via relay), but blew its fuses when I tried it with my Croft Micro II (valve regulated, so should have been ok, but I’ve upped smoothing capacitance in it significantly) with power amp on first.

Kind of interested if I can detect an SQ difference with and without :)

Volume pot arrived yesterday, so I can slowly start build next week.

Richard
 
You must log in or register to reply here.


advertisement


Back
Top