|
|
Post by julesd68 on Nov 19, 2016 19:18:07 GMT
Does anyone have much experience of 5AR4 / GZ34 tubes?
I've not used these before but my new preamp has one and I'd like to try a few different types with it ...
At the moment there is a Sovtek in it which I doubt will be too hard to improve on.
I've also got a later production Mullard to try.
If anyone has any recommendations for other NOS varieties I would be very interested ...
|
|
|
|
Post by pre65 on Nov 19, 2016 19:33:56 GMT
Just be aware, that GZ series rectifiers are indirectly heated and have a warm up time before full HT is delivered.
There are "other" types that will physically fit, and work, but are directly heated and have much shorter warm up time. Also check data sheets for heater current as that can vary as well.
Also, the HT available can vary between different types.
I'd expect a proper Mullard to be the best bet. 
|
|
|
|
Post by julesd68 on Nov 19, 2016 20:31:16 GMT
Thanks for that - what kind of warm up time are we talking about with the GZ34?
Problem with the Mullards is that there are so many different variants it is difficult to know what is the real deal and what is just mediocre and still expensive as I found out when trying a variety of Mullard 12AX7 some time ago ...
|
|
|
|
Post by pre65 on Nov 19, 2016 20:59:10 GMT
I can't find the definitive answer at the moment, but from memory it was about 5 - 7 seconds. If the amplifying valves are indirectly heated then I would suggest the rectifier should be as well. Another thing, the valve data sheet will probably list a minimum anode resistance, which would be 50% (half) of the transformer secondary resistance-ie centre tap to each anode. It will vary according to the actual HT achieved. If the actual figure is lower than necessary then resistance should be added. Arcing of the rectifier when switched on can be one symptom. As an aside, on most of my home build valve amps I use 6D22s damper diodes (in pairs) as they have a warm up time of about 30 secs. Here is a GZ34 data sheet, the resistance mentioned is Rt frank.yueksel.org/sheets/030/g/GZ34.pdf |
|
|
|
Post by julesd68 on Nov 20, 2016 12:02:40 GMT
I'm afraid that's all going way above my head Philip!
I'll have a look to see if there are any decent Mullard alternatives...
|
|
|
|
Post by MartinT on Nov 20, 2016 13:16:13 GMT
I hate to say it, but I upgraded the Leak Stereo 20 I had many years ago with a RATA solid state module in place of the GZ34 (heresy)! It sounded better to my ears, less soft and with more detail.
|
|
|
|
Post by julesd68 on Nov 20, 2016 13:45:01 GMT
I hope I can sleep tonight Martin, I will try and forget you said that ...   |
|
|
|
Post by MartinT on Nov 20, 2016 14:20:19 GMT
|
|
|
|
Post by ant on Nov 20, 2016 17:46:06 GMT
Another thing to consider is the voltage drop with these. I usually use gz 34s in place of ones like the 5u4g type as they drop somewhere in the region of 30v.other will drop more and an ss rectifier significantly less.
If for example a bit of kit will be run somewhere around 300v ish this is a 10% drop that the designer will have designed around. Modifying an old amp with an ss bridge could push the operating voltage higher than intended, or swapping in a different type a lower v.
Consider that the valves if cathode biased (any other bias type will still be affected) will have their operating current set by a resistor and capacitor between the cathode of the valve and 0v, a difference in the ht voltage will result in a higher or lower quiescent current, which is the amount of current the amp idles at.
This will change the bias point so the valve could be working outside the design centre for the circuit.
If it's a pentode this will also affect the screen voltage which is a fine balancing act at the best of times.
Now 99% of amps designed in the UK for a gz34 will be probably based on mullard data, so a mullard is a good bet but will be dear. I have found that jj ones can have problems with arcing and sovtek ones are reasonably tough.
Exotic ones I wouldn't waste cash on.
Bear in mind that it is nowhere near the signal path and is isolated from the ht at the anodes of the valves by the set of caps usually after it. so in my opinion, one rectifier sounding different to another is due to electrical changes in the current and and voltage supplied to the amp.
Be aware also that replacing with an ss rectifier can put stresses on the power supply caps as they will get instant voltage that doesn't build up over a couple of seconds, but is just there instantly. Old amps with old caps might not take kindly to this
Cheers ant
|
|
|
|
Post by julesd68 on Nov 20, 2016 18:14:34 GMT
Well I understood the second half of that!
With those words in mind, I certainly won’t order another 34 until I can work out if my Mullard makes any difference to the presentation.
Thanks for your input.
|
|
|
|
Post by ChrisB on Nov 20, 2016 18:28:51 GMT
About 20 years ago, I bought several old unused Mullards for 17 quid each and I see they're going for in the region of a hundred now. I only have two left. One is in my Radford and the other, I hope will be all the GZ34s I will ever need!
|
|
|
|
Post by ant on Nov 20, 2016 18:37:27 GMT
Just an additional thought, different brands of 34 are a safe bet either way, they will all have electrical characteristics within a certain range, and some will be manufactured to better tolerances than others. A 34 is a 34 is a 34. in other words, replacing like for like will not harm the amp. One will sound better or worse to your ears than another. Valves are pretty forgiving, a 5% tolerance in ht voltage or current isn't going to bother it too much. 10% plus depending on the bias point for the amp could be pushing it. Id just say stick to the type if you intend to tube roll, and if you spot something that proports to be an alternative to it, check the dataset first, or get someone to look at it who can interpret the data correctly before trying
Also don't be too sniffy about the ex Soviet stuff, most of it is military grade stuff so can be pretty good even though it's still pretty cheap
Cheers ant
|
|
