Kat34 component replacement

[FoxJT]

Hi,

I am upgrading my KaT34 (was a KiT34) and have just noticed that at least two of the EL34 cathode bypass capacitors (100uF 63V) have "blown"; i.e. their bottoms are bulging. A cause or effect of EL34s going very red!

I intend replacing them with Maplin VH39N. However, before I source them, are there any recommendations for a better replacement?

Also, there had been talk a long time ago of perhaps changing the cathode resistor from 390R to 470R . Is this worth while?
Ta.

Jeremy.

[ian_r]

I'm sure you could spend much more, but you might also consider starting with (for a similar price to Maplin)

http://www.cricklewoodelectronics.co...cat=156&page=1

I have found that cathode bypass capacitors (which are in the direct signal path) have a definite effect on audio quality.

[Richard]

Yes Ian's rec should be good, or if Maplin is convenient, the ones you suggest are good quality.

Consider also DT62S 100uF 50v which are 105 degree rated and will be low impedance as they're specced for SMPS so might sound a tad better than ordinary caps.

My thoughts are that the amp is a few years old, the original caps were basic quality, and have dried out.

High temp ones will last longer and are often low impedance due to the change in use of today's typical power supplies so you should get better service, and possibly better sound, from them. From the Maplin site,

"High Temperature Radial Electrolytic Capacitors



A range of high
temperature, miniature radial electrolytics offering excellent reliability and long life - up to 2000 hours at 105°C. The capacitors are ideal for use in the output circuits of switching power supplies and other high temperature environments such as monitors and automotive applications. At lower operating temperatures, life expectancy is well over 10 to 15 years."

Re changing the cathode resistors. The amp is biased for full class A I think. HT is a little lower than the popular 5-20 OP stage and the lower value resistors bias the valves for a little more current to hold it in A. This gives a plate disippation of 26W which is about max for el34.

You could change to 470R which would give the valves a slightly easier time and there would probably be little difference in the sound. It should be fine as it is though, especially if you're fitting those rect balancing resistors which will bring HT down a little.

When it's done check the valve plates for any red patches in a very dark room. That will give an indication of any overheating problem. Otherwise it was probably just old age.

rich

[FoxJT]

Thanks for the suggestions and tips, guys. Old age is probably the reason as it is about 4 1/2 years old. My old age may be a factor as well

Sadly, I don't have a Maplin near me (although I do have an RS near where I work in Crawley), so it is an annoyance when buying a small number of components that postage doubles the cost

Rich, while I am waving my soldering iron around, would you recommend that I apply the same thinking about the caps to the smoothing caps in the power supply? And also go the whole hog and change the electrolytics (1000uF 10V) in the ECF80 cathode.

Adding the rec balancing resistors has dropped the DC volts nicely. The rec cathode volts dropped from 435 to 422; still higher than the 408V in the diagram, but my AC here is high: about 243VAC.

Cheers,

Jeremy.

[acorn]

Yes I would go along with Richards suggestion with regards to the Maplin high temp caps go for the 100uF at 100vdw or 50 uF if you want a faster amp and change the 1000uF while your at it and you could try a 470uF in this position for a cleaner response with the same type, thats what I use in my amp.

With regards to the bias resistor I would go for the 470R type with a 422 Ht rail as Richard said it would pro-long the life of the valves, haveing said that I have a 380 HT rail and 390R bias resistors and have had know trouble so far with my output valves I always try and under run them rather than over run them and the amp sounds fine.


Acorn

[Phil Y]

Hello Jeremy,
Just as another opinion, i rather like Elna Starget caps, i think they sound better than the low impedance ones from Maplin. They are available from RS which sounds handy for you. The RS number for the 100u 50v is 215-5764.

Cheers, Phil.

[Richard]

Originally Posted by FoxJT

Thanks for the suggestions and tips, guys. Old age is probably the reason as it is about 4 1/2 years old. My old age may be a factor as well

Sadly, I don't have a Maplin near me (although I do have an RS near where I work in Crawley), so it is an annoyance when buying a small number of components that postage doubles the cost

Rich, while I am waving my soldering iron around, would you recommend that I apply the same thinking about the caps to the smoothing caps in the power supply? And also go the whole hog and change the electrolytics (1000uF 10V) in the ECF80 cathode.

Adding the rec balancing resistors has dropped the DC volts nicely. The rec cathode volts dropped from 435 to 422; still higher than the 408V in the diagram, but my AC here is high: about 243VAC.

Cheers,

Jeremy.

Hi Jeremy,

It wouldn't hurt if you are looking to keep the amp and have reliable listening from it. Electrolytics are probably the most unreliable of all the components (except valves perhaps and they usually just fade rather than go catastrophic).

If you can shop at RS the choice is much greater. Look at the Evox Rifa range of electros as used by WD for some apps. They have a very long life (born from supplying the telecoms industry I think) and spec for spec usually have lower esr which is good for sound. Certainly I've done direct comparisons swaping them into power supplies and they make a good difference to cheap brands.

All said though, psu caps will usually give plenty of notice by causing humm if they are packing up, and small signal valve bypass' don't usually explode if they go short as the valve can't pass enough current. It would damage the valve though.

Rich

[FoxJT]

Rich, Guys,

Thanks for the tips. Plenty of food for thought. I am currently "house bound" recovering from an eye op so I won't be able to get to RS for another week or so, but I will check them out on the web and visit once mobile again.

In the mean time I do have some replacements to check out the amp. I am happy with it currently (and now voltly ) but I would love to have a go with the WD88VA sometime soon. Start saving .

Jeremy.

[acorn]

Hi Jeremy

Another point to consider is to by - pass your output valves cathode caps with .68uF PXX Poly type from World Designs which I have done and sound good too, this will result in a faster and more decisive response.

Acorn

[FoxJT]

Acorn,

It always sounds rather bizarre putting a very small value cap in parallel with a very large one; 100uF||.68uF, especially when you consider that the tolerances can be +/- 20%! What possible difference could it make!

However, I assume that we should take into account the ESR + ESL and the EPR (leakage resistance) for each component to give the complete picture.

Checking with MJ, paralleling capacitors of small values can tend to make the composite component approximate a perfect capacitor, although he does go on to say that bypassing used to be more important but cap design has much improved.

I'll have a go at that later after current mods have been listened to. As they say, one thing at a time!

Cheers,

Jeremy.