KAT6550 finally built but some help appreciated

[timsymons]

The construction of my KAT6550 is now complete. The story so far:

I decided from the off to make some changes to the components and construction of the amp, which I'll detail shortly. I started off mounting the hardware, making a gasket for the mains Tx from rubberised cork, with another under the choke. Next I did the heater wiring. Here I decided to depart slightly from the published layout as I wasn't too keen on running the heater wiring under the channel tag strips. Instead I ran the wiring from the PSU tagstrip back towards the choke, then taking it out to near the chassis edge, and running it in to each valve base, then back out, as recommended in Morgan Jones' book Building Valve Amplifiers. Taking to heart his remark that, with good heater wiring, even a phono stage can be run with AC heaters, and as I didn't have the components to implement the heater mod with the extra resistor and capacitor, I decided to run the ECF80s' heaters on AC. I also decoupled both legs of the heater wiring to the chassis at each valve with a 10 nF ceramic capacitor (again, as per the Morgan Jones book), having first removed the paint round the screw hole to make a good connection. Having put in all the heater wiring, I tested it - so far, so good.

Perhaps the biggest component changes were in the PSU. I replaced C1/C2 with a Mundorf TubeCap 47 uF@600V and C3/C4 with half of a Black Gate WKZ 100+100 uF @ 500V, with the other half acting as C5.

As the tagstrip layout in the instructions didn't accommodate the RC network R47/C23, I worked out a new layout to include them. This didn't leave any room for the 6550 cathode bypass caps. I then hit on the idea of mouting the cathode resistor and bypass cap directly on the valve bases, using pin 6 (nc) as the 0V end. All the electrolytics in the audio circuits are BG standard types. The coupling caps are Mundorf Silver/Gold types.

To test, I connected up a 4 ohm resistor and one of my old Kef C60s to each channel with an old CD player for input. Much relief when it worked first time.

On putting it in my main system, I was very impressed with the sound - very clear and open with bags of detail. My only problem is noise. In the L channel there's a miniscule amount of hum, only audible with your ear right by the speaker, inaudible from more than a few inches. Slightly more in the R channel, which I'd like to clear if I can. As the L is so hum-free, I'm guessing it's a wiring thing, so maybe moving some wires, particularly the feedback wire might help, although I haven't tried it yet. More problematic is a noticeable hiss from both channels which, while you don't notice it when music is playing, is audible when it stops and I'm anxious to fix. Both the hiss and such hum as there is disappear entirely when you take the ECF80s out, so it it must be coming from the input circuitry. As the hiss is identical in both sides, I'm guessing it might be PSU-related. One thought was that R5 might be noisy. It's dissipating 1.85 W, which should be fine for a 3W component, but it is common to the power supply of both ECF80s. I've tried different rectifier valves and ECF80s with no discernible change. I've attached some pics and listed some voltages below. The voltages all look fine, if a bit high, so it's difficult to know where the hiss might be coming from. Any help very gratefully received.

Many thanks,

Tim

Test voltages

Mains 243 V

HT before choke 498 V, after choke (point A) 487 V. Point B 385 V.

ECF80 (L, R)
pin 7 1.71 V, 1.75 V
pin 3 94 V, 93.9 V
pin 6 144 V, 139.7 V
pin 8 6.7 V, 6.6 V
pin 1 136 V, 132.1 V

All 6550 cathode voltages 46.8 - 47.2 V

[Richard]

Hi Tim

Very nice. I feel the noise/hum will be due to going ac for the ecf80 heaters and that even using the standard rect and cap it will be very quiet. Perhaps the triode/pentode ecf80 will be noisier than the types MJ would use for a phono amp. I'd fit dc and if it cures it then at least you'll know and will be able to consider improving further to the schottky pi arrangement in the faq.

Rich

[JerryT]

Hi

I've used KAT6550 for several years. Hum problems are usually to do with wiring and can be altered easily and removed with persistence! However they also seem to return with the slightest alteration in components, peripherals or even the weather (perhaps mains quality, local DC - who knows) .

Hiss problems I've not met . I'm currently using my amp to drive Fostex 206E s in a quarter wave pipe (96 dB/W) and there isn't any hiss. If you have significant hiss I would look closely at the resistors in the early signal path and replace them with higher wattage and/or lower noise types.

Early on I replaced the supplied EI ECF80s with Mullard E80CFs. This certainly improved clarity/reduced noise and reduced the 'greyness'. Changing cathode bypass caps to Black gate afterwards improved the same effect. Perhaps these two changes also reduced the HF noise floor, although that wasn't the effect that I noticed at the time.

Don't be deflected. The KAT 6550 is a very good amp that repays a little time on tuning/tweaking. Enjoy.

Jerry

[timsymons]

Hi Richard,

Thanks for your reply. I'm going to get the parts for the DC heater mod etc. and will report back when it's done.

Tim

[timsymons]

Well, I've installed the DC heater circuit with pi filter with frankly mixed results. I used 4 MBR360 diodes with 4700 uF across the output, followed by a 1R resistor, another 4700 uF, then 0.68R and a third 4700 uF cap. I used this configuration principally because Maplin didn't have a 1.5R 3W resistor...

Unfortunately, the hiss is still there. It may be very slightly reduced, but it's marginal. It seems to be quite broad spectrum, by which I mean you can hear a component of it in all 3 drivers, although it's probably loudest in the mid driver.

What has certainly increased is mechanical noise from the mains Tx, presumably due to the ripple current of the DC circuit. This was one of the reasons I went with AC in the first place, having read posts over on the old forum about mechanical noise with DC heaters. Pressing my hand on top of the transformer does remove it, so maybe a tightening of the mounting bolts would help, but it was silent with AC heaters. Grrr...

I don't think I'm any further forward - given the Tx buzz, it's sort of gone backwards. Does anyone have any suggestions please?

Many thanks,

Tim

[Mike G]

Hi Tim,

I have exactly the same mechanical hum from the mains transformer on my KAT6550 too, and like you this is greatly reduced if a hand is pressed on top!

Unfortunantley this makes listening rather awkward!!

Luckily the hum is not that bad at all so i've never felt the need to investigate further, however I have tried a cork gasket underneath the Tx, which reduced it very slightly. Tightening the Tx down didn't really help, so I tried the opposite approach, and loosened off the bolts - effectively decoupling the Tx slightly - this did help!

Even gently pressure on the Tx cap, seems to reduce the hum, which i'm now sure is amplified by the resonant effect of the Tx cover. I have considered exploring the possibility of placing some form of damping material, inside the Tx cap, - although at present i'm not sure what material would be best, as the Tx runs pretty hot!

It maybe worth reading this thread - regarding the DC blocker.

http://www.world-designs.co.uk/forum...hlight=blocker

Regards

Mike.

[timsymons]

Hi Mike,

Thanks for your reply. Unfortunately, the buzz on the Tx is pretty noticeable, even from a few feet away, so something needs to be done, and I've already fitted a cork gasket.

I'm going to get the 'scope out tomorrow and see if I can't find this hiss. My suspicion is that R5 is noisy. I have some resistors is my spares box I can cobble together to make a replacement to test that theory if needs be.

One thing that slightly bothers me is that the HT secondary is high - the circuit diagram gives 425-0-425 V, while mine gives 450-0-450, which obviously gives high(er) DC voltages. As others have noticed, the mains Tx does run quite warm.

The one good thing about this (so far!) is that I feel vindicated in going for AC heaters on the ECF80s.

We'll see.

Cheers,

Tim

[Richard]

Hi Tim

The hiss is puzzling as I've had a Kit6550 in my system and it was completely silent. In fact I can't remember any WAD valve amp hissing and have had ss amps with more hiss.

I'd try separating hiss from hum as they may be different issues and dc should have improved the electrical hum component of the noise.

Try shorting the inputs and see if it goes - suggesting something on the input causing it. If not you may have a noisey resistor. Is it common to both channels, maybe a power supply resistor, or just one channel, a circuit resistor perhaps on the front end. Have you used Holcos, they are fragile and can go noisey.

Are your speakers very sensitive?

Mechanical TX noise is a different thing of course and has been reported here, though again, the one I used was quiet. Maybe mains varies, I may be lucky. Try the different mountings suggested and try to decide if it's a buzz from loose windings/lams (return it for re-impregnating with varnish) or a mounting issue. It may be exciting a resonance in something else - or could it be the choke - they play up sometimes.

[timsymons]

Made some good progress this morning (day off!) Had the amp on its side with the baseplate off to do some testing and the mechanical noise had gone - turns out the ripple current was exciting a resonance in the baseplate. I've got some foam-backed draught-excluding tape somewhere so lining the baseplate with that around the edge should sort it out.

Had the scope out on the PSU and it seemed that there was more noise on the HT after R5 than before, suggesting R5 was the problem, although I did have to turn the vertical sensitivity right up. Then I tried shorting the inputs and it turns out I've been looking in the wrong place all along - both channels were as silent as the proverbial grave. There was the slightest hiss, but I had to get my ear right up to the speaker cone to hear it, and even then it was very faint. So it looks like my Conrad-Johnson PV6 is the culprit. What threw me off the scent was that with my previous power amp arrangement (the power amp section of an aging Technics SU-V707) I didn't hear any hiss or hum at all. The way I had it set up was that the PV6 was fed into the graphic EQ return sockets, which are post input selection, but pre volume control. The Technics was set on 2 and the volume control on the PV6 was the 'real' volume knob. Bit Heath Robinson, but it worked.

In answer to your question, Richard, the speakers are pretty sensitive. With the PV6 feeding the 6550, 2 on the volume control is plenty loud enough. I was thinking of trying to reduce the output of the PV6 with a larger series resistor, but now I've found the hiss etc, I may try other options. Maybe a passive pre with valve phono stage is the way to go.

Anyway, many thanks to all those that replied. At least I can stop worrying about it now ...

Tim

[Mike G]

Hi Tim,

That's good news! - I forgot to mention the base plate of the KAT6550! As the screws are in the corners, it does tend to vibrate, I added extra screws to hold it firm.

Mike.