Krell KSL Phono Board MC Issues

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
By chance anyone have a schematic for a Krell KSL MM/MC phono board?

Trying to run down a hum or buzz issue on the left channel only when set to MC. This is with or without a turntable connected. For now just the R and L channel outputs are connected to amp. MM is quiet on both channels; I just get a slight hiss at each speaker when cranked to full volume.

When set to MC, I get this hum or buzz that becomes audible around 10 o’clock from the left channel and increases with added volume. Unfortunately I listen at 12-2 with my low output MC. I just get a hiss from the right channel at full volume set to MC. Left channel noise is present no matter what load setting is used. I’ve even swapped R and L outputs and the noise moves to the right speaker.

So far I cleaned the old nasty flux from bottom of board with alcohol and tapped/wiggled components. No change.

An externally hosted image should be here but it was not working when we last tested it.
 
I recently rebuilt a KSP 7B which has a broadly similar internal phonostage, MM/MC options with various gain and loading settings.

From what you say it sound slike the RIAA and MM gain stages are fine, on the KSP the MC stage was an additional gain stage that was switched in before the MM stage.

On my phonostage I was 6db down on one channel and suffering from noise, it was a blown transistor in the first gain stage that had gone. I replaced all the Rhode capacitors as well just to be safe, (made no difference ;-)

If you have different noise from one channel to the next then that might help you find at least one fault. Measure all the resistors and make sure the bad channel matches the good. Do the same for the small signal capacitors and then measure all the transistors to make sure none of them have shorted, an open short will show up easily as differing junction resistance measured across the legs.

The red Rhode caps will likely be fine unless the entire amp has been fried- they do run warm. The silver caps however might be well past their sell by date. I would replace these before getting too deep in searching for difficult to find faults.

A word of warning though. The pcb's are made of hopeless fibreglass, they do not like being reworked and the vias will pull through easily as they are very tight to the legs of the larger components. If you are removing a component make sure you snip the body off first, so you just remove the component leg, you really do not want to try and remove a large value capacitor from the pcb as it will suck up heat and destroy the pcb/via.

I used a 45 watt iron and had to be extremely careful. i am a very delicate solderer, so treat your pcb very carefully, the Krell is the worst pcb I have worked on.

krell ksp-7b - a set on Flickr
 
Thanks for the reply and tips on the board. I had desoldered one end of the two diodes on the left channel and they tested OK. I have a Weller solder station and good solder sucker. I can also borrow a Hako 808 desoldere gun from a frined if need be.

I started measuring voltages and found the bias for the transistors at the right end of the board near the pot are .75v higher on the left channel than the right channel. The left channel is also outputting near 1vDC set to MC while the right channel output is 0v DC. I'll keep measuring.

The silver caps on the phono board are Sprague. I was wondering if they were the pricey Tantalum? Part number is H8016 or 9108H, depending on how you view it. I'm no tech but have read that some think Tantalum caps are not worth it? I couldn't afford to go that route at $50 each.
 
Last edited:
There's nothing really remarkable about Tantalums. I have a fair bit of experience replacing them in Naim equipment and trying NOS Sprague wet tants in my phonostage coupling positions. I prefer poly caps in all cases instances- though this isn't an option for you obviously.

1vDC on the output? Should be easy to track down with a scope. Best of luck.
 
Member
Joined 2005
Paid Member
Several resistors in the first row next to the dip switches might not be stuffed in the right position - probably they are loading resistors, it will still work , but not load the cartridge equally on each channel.
Or Maybe they are allright ands it's just a difference between channels for board layout.

this isn't likely the problem you have - thats likely a bad cap or transistor I suspect.
 
Last edited:
If the trouble is only with the MC setting, follow the traces from the MC/MM switch back.
You will probably find some transistors and PS caps and resistors.
Have a look at those.
Everything downstream of the switch is likely okay if the MM setting is quiet.
 
Thanks for all the input/help.

1vDC on the output? Should be easy to track down with a scope. Best of luck.

Unfortunately my old Tektronix 466 has flat lined on me. It won't display a trace anymore; just a horizontal line appears on the display. Even when connected direct (no probe) to the square wave test port on rear of unit. So all I have is my trusty Fluke MM.

I appreciate your input on the caps. I would entertain a better than plain replacement cap (even recap the main board) once I am past the MC issue. I have de-soldered one end of each cap and they all measure close to value (using Fluke MM set to capacitance).



Several resistors in the first row next to the dip switches ....

I had noticed that also. It looked odd so I compared it with some web photos of another KSL phono board and found they are the same.

If the trouble is only with the MC setting, follow the traces from the MC/MM switch back.

I tried that and it isn't as obvious as I thought it would be.:confused:


So far with board removed I compared readings of resistors/caps from side to side and readings are same/similar. One end of each didoe was de-soldered to test. Some resistors start climbing in value with MM attached, but initial probed reading starts out similar for each side/channel. Some of the resistors with climbing values measured in circuit I actually de-soldered one end and measured them.

I couldn’t locate any open or shorted transistors, but have isolated a powered portion of the board where I believe the issue to be. I have a nasty head cold which makes it harder to wrap my head around what this all means. It is at the south end of the board where pos and neg power enter through metal standoffs (allen head bolts above caps) from main board. Right side allen head is neg and left is pos input power (both 20vdc).

If I bias Q15 of left channel to match Q38 of right channel, that -.072v number gets larger, like .114 or higher. That increases base of Q20 where the right channel Q43 is -.005. With resistors measuring similar I am leaning toward transistor problems, since some transistors feed others. That -.072 reading is also very similar to the DC on the left channel output.

The transistors facing left are MPS 8599 PNP, the ones facing right are MPS 8099 NPN. MPS8599/8099 are now MPSA56, MPSA06. I see some MPS8599 on ebay but.....maybe not enough, not sure of you can trust the source, etc. Or is there an equivalent/different parts to use? Change all transistors in this area?
 

Attachments

  • phono_1.jpg
    phono_1.jpg
    345.2 KB · Views: 644
Last edited:
Well duh on my part, thanks for that info. The 30D was staring me in the face and I went with the H number. The H number turned up a tant reference.

Lots of duh here lately. This board has some J309 jfets among the transistors. Eight up by the dipswitch end, and two above each outer bias pot (Q13, Q14, Q36, Q37). One of the left channel jfets has some different readings than the other on the right channel. I also learnt that the adjustable pots at the middle of board must be for nulling DC. Nulling the DC on the left channel with the pot helps reduce the hum, but it is still present.

Luckily my DAC enclosure arrived so I can leave this phono board alone for a while.
 
Last edited:
Thanks again for all the tips. I finished my DAC and got back to the phono board. Some progress made but more needs to be done. I replaced the old Sprague caps on the phono board with new 105C Vishay BC's. No change in left channel noise, but all caps measure similar uf now (measured with Fluke set to uf) where the old ones varied from 5-20%. I’ll have all new caps on the way this week for the pre amp.

I got some NOS J309 jfets. I used a 9 volt battery/multimeter set to mA to match them. (Neg of DMM to Drain, Pos of DMM to B+, jumper B- across G and S). They ranged from 2.04mA - 2.07mA. The second of the two removed jfets measured 2.08 (the first is my flaky one). I installed two new jfets in the left channel and powered it up.

Now I never touched the right channel bias pot and I would get different readings from time to time. It usually measured ~.370, every now and then .440. I just ever so slightly turned the adjustment and it snapped to .608. Turning right or left now gradually increases or decreases bias. It would take many turns to get to .369 so I am guessing ~.600 is where it should be.

IF I attempt to adjust the left channel bias over .369 to .600, that left channel circuit that measures .072 increases which means more DC to null out at the left channel output. (With the new jfets, that .072 dropped to .026). I left the left channel bias at .369 and get much less noise out of the left channel now. I'm thinking I'll have to replace some or all of the transistors like sq225917 did for a total fix.

If I did replace all transistors (16 - MPS8599, 18 - MPS8099, no longer manufactured) should I procure some NOS MPS8099/8599 transistors or their replacements (MPSA56/MPSA06)?
 
Last edited:
Some closure here. On board testing wasn't that reliable so I removed every transistor and jfet from the left channel and tested them; none failed. Out of frustration I reassembled and adjusted bias (pots at sides of board) until I got zero DC in that one what I call bias circuit (cause I don't know what else to call it) in each channel. That bias circuit sets bias for a transistor by the null DC pot (by center of board) and sets transistor E voltage at end of board (follow the .072 measurements in previous pic; the R channel bias circuit was .006 and quiet). While I could null that DC at the output, the left channel hum was still there.

I don't know if I did it right but the left channel is quiet now. So the right channel ends up at -.612, and the left channel at -.560, zero DC in that ‘bias circuit’ and all voltages are pretty square from side to side. L and R volume sound identical, vocals centered. Some test tone tracks and output voltage tests could confirm all is well. All due to dirty or sustained transfer spots in the pots.
 

Attachments

  • kslphono_recapped.jpg
    kslphono_recapped.jpg
    259.1 KB · Views: 438
Last edited:
Finally got around to recapping the KSL main board and power supply. Some measured close on the Fluke set to capacitance, some were up to 27% over spec. I don't have an ESR tester. A friend that I helped once helped me with 22 United Chemi Con KME's. Digi key supplied two 4700uf UCC KMH's for the power supply, and four really long life 470uf UCC LXY's for the hot spot between the sinks.

He parted with some Siemens Styroflex polystyrenes to replace the four .033uf caps on the phono board. I don't know if they are part of the RIAA filter or some signal feedback loop? I'll look for some .0016uf's to replace the two remaining poly's. It still works and sounds pretty good! Thanks again for all the help and tips.
 

Attachments

  • kslrecap_1.jpg
    kslrecap_1.jpg
    165.1 KB · Views: 382
  • kslrecap_2.jpg
    kslrecap_2.jpg
    136 KB · Views: 380
  • kslrecap_4.jpg
    kslrecap_4.jpg
    166.2 KB · Views: 469
  • kslphono_recap_2.jpg
    kslphono_recap_2.jpg
    101.5 KB · Views: 326
Sorry, I'm not a tech or that smart kamis.:confused: I don't have a schematic and have not been able to get one. Due to the multi layer board it is difficult to follow the circuit....for me anyway.

EDIT: An added note, I have a friend that works in the electronics business. For what its worth he said the parts count per gain stage seems high for a balanced discrete design but would need to see a schematic.
 
Last edited:
Sorry, I'm not a tech or that smart kamis.:confused: I don't have a schematic and have not been able to get one. Due to the multi layer board it is difficult to follow the circuit....for me anyway.

EDIT: An added note, I have a friend that works in the electronics business. For what its worth he said the parts count per gain stage seems high for a balanced discrete design but would need to see a schematic.
Boards are double-sided, not multi layer. You could read down transistor markings on 8 transistors in row near DIP switches. I would be grateful.
 
Boards are double-sided, not multi layer. You could read down transistor markings on 8 transistors in row near DIP switches. I would be grateful.

Well it looks multilayer to me as 1) there are traces visible with a strong backlight that are not on either side of board. 2) There is a seam at the side of board. I don't have close up capability to show that well enough.

Here are some pics, best I can do.

http://i417.photobucket.com/albums/pp260/scompracer/Krell KSL/close_1_zps7f874d8a.jpg

http://i417.photobucket.com/albums/pp260/scompracer/Krell KSL/close_2_zps25002355.jpg

http://i417.photobucket.com/albums/pp260/scompracer/Krell KSL/close_3_zpsa8948606.jpg

http://i417.photobucket.com/albums/pp260/scompracer/Krell KSL/close_4_zpsfe5419e4.jpg
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.