Harman Kardon HK 330B recap, restore and upgrade BOM part list (kinda)

Hi,

My amp lost sound in "speaker system 1", FM Stereo was noisier than FM, so I have finally pulled the plug and upgraded my HK330B with the first BOM list, adding Rectifier main and Rectifier SAB parts from the extended BOM. The only difference is that I used a 3300uF big cap instead of 4700. Also, my MPX board has the 50-75 switch that another user showed in his pic.

It powers on without any thump now, the AUX works well on both speaker systems. I've lost the FM and AM, which came back after I touched on of the resistors on Rectifier board. Weird, but probably my fault.
Still some background hiss in FM stereo but volume is much better overall. I think I'm good for a while.

Thanks!

PS. Forgot to add, I'm running the old Admiral Tunnel Reflex with it, with its Fostex speakers replaced with Radioshack equivalent from the 80s, since the original blew their suspension :)
 
Last edited:
Good morning
on the RECTIDIER MAIN board can I replace diode D2 with a 1A 400V diode?
 

Attachments

  • AA.png
    AA.png
    165.7 KB · Views: 16
Dlucy, your lists for the HK330B-1 are invaluable. What is the equivalent for the 2sc460? This is from another thread.. I’ve tried the KSC945CYTA (with center base) to no avail. I’m a newbie at this and loving it.
 
So @dlucy, what are your tips for getting into the lamp shround to replace the dial lamps on a 330b? One of the dial cord pullys is connected to the shroud making it difficult. I'm not sure you can get in through the front like on a 330c.
 
Namlook said:
So @dlucy, what are your tips for getting into the lamp shround to replace the dial lamps on a 330b? One of the dial cord pullys is connected to the shroud making it difficult. I'm not sure you can get in through the front like on a 330c.
Click to expand...
I just had to replace the lamps in one yesterday.

What I did was to attach a wire around both dial cord sections, then pulled them off the pulley towards the corner of the face plate. I then wrapped the wire around the corner of the face plate and back towards the cords to keep them tight. You can wrap the wire around itself or on the cords to hold them out of the way. I wrapped the other end of the wire back onto itself to make replacing the cord onto the pulley easier.

Be sure to note the way the cords go so they get replaced on the pulley right.
 
Last edited:
I have specific questions about my 330B restoration project, but I'll post that separately. But I do have some generic BOM thoughts.
I'm basing my information on my series B-1 with serial number ending in A.
  1. All of the medium-large Elna 220uF/16V capacitors can be replaced with 330uF or 470uF capacitors, because they are all rail decoupling capacitors in this design. On my unit, most boards have 11~12V rails so I put 470uF/16V caps, still 25% voltage margin. But my control board runs on a ~16.5V rail, so I recommend increasing C519 and C520 to 25V+. Factory manual says the transistors on that sub-board see 17.2V, I think this 25V+ the right recommendation.
  2. The "Echo Japan" 220uF/6.3V capacitors C403/C404 are emitter bypass capacitors so keep at 220uF--changing the uF will change the transistor behavior.
  3. I have a clean unit, and I don't normally try to preemptively replace trimpots on restorations, but my VR401 trimpot broke apart (pieces went flying!) as soon as I touched it to perform the factory calibrations. So, replacing the trimpots seem mandatory if you want to replace the transistors, since you really should recalibrate after transistor replacements.
  4. Consider adding fuses to your BOM. My unit didn't work properly until I re-seated my fuses and sprayed deoxit inside the fuse holder.
  5. As far as I can see, none of the 2SC458 transistors are matched pairs, so the replacements don't need precise hFE matching. Maybe some coarse matching to maintain L/R channel balance. Q405 and Q406 are adjusted during the VR403/404 trimpots, so transistor matching is not critical there either. Can anyone confirm or correct me?
  6. For R3, 100Ohm 2W, my alternate BOM recommendation is 100Ohm 4W 100V wire wound. There is plenty of space for the larger 4W. Suggest paying attention to voltage rating since it is in the 53VDC path. This is the only part of my unit where the factory circuit board was a singed/discolored from the 2W giving off too much heat. I'm recommending wire wound because it's on DC rail and is expected to face surge current. I'd love to learn why metal-film was recommended by dlucy. Same question for R1/R2/R4, all DC power rail resistors, not audio path, so I cautiously suggest alternate BOM recommendation of using wire wound. But I'm probably missing something obvious. (I forgot that wire wound gets expensive for larger resistances.)
  7. As others have already confirmed, LED replacements for the fuse-style bulbs is easy but does make the unit a lot brighter.
Please feel free to correct where I've gotten it wrong. Thanks!
 
Last edited:
dlucy said:
Gimme a day or two and I'll assemble some photos and description on the trimpots. Replacing those won't make any audible difference, BUT Ive had the trimpots from more than two units literally fall apart as I tried to adjust them. That is a BIG risk to the unit if power is on. But replacing the trimpots are definitely optional.
Click to expand...
As of 2026, whatever trimpots are still in the wild definitely fall apart if you touch them. So I'd say they are not optional if you want to make the factory calibrations. One of mine spat a piece somewhere across the room.

Measuring my factory 3K trimpot on my multimeter, I got a non-linear response between 1kOhm and 4.5kOhm, and it jumps to different values as I turn it. I suppose I can save it with faderlube.

For restoration, it feels mandatory to replace these for anyone breathing around the circuit. However, the correct equivalent to mimic factory adjustment range would be [1kOhm fixed resistor + 3.3k trimpot] in series, where values below 1kOhm is not allowed. The Bourns trimpot adjust down to 1% (just 30 Ohms for 3K max) so that could be dangerous to the circuit if someone is careless.

OR maybe the original trimpots used to reach 0Ohms and mine is just old...
 
Last edited:
Man, my partially restored 330b sounds absolutely wonderful!
It was in original condition (letter A in the serial, still made in Japan). Right channel output cap was shot, fully working when replaced. Sounded a bit muffled. Most, if not all, 10uf caps had high esr. Bigger caps were allright, some still to be replaced. I used films for 1ufs when they were between the stages.

There was nothing wrong with the original outputs, but they looked so shot I replaced them. I had mje15032s in stock, so put them in. Scope showed oscillation at medium volumes. Replacing the drivers with c2383s and a1013s fixed it. Ended up replacing rest of the trs too (mostly c1845s), and the flimsy looking trimpots. Refurbished the power supply too, but original mfc is still in.

It sounds a tad warmer than 730, and if there's a difference in detail retrieval, it is pretty insignificant. Well pleased! (no burn-in time needed btw).
 
After installing new mfc (6800uf), new output caps (1000uf UKW) and low leakage 10uf Rubycons in selected places there is a subtle, yet notable improvement in clarity (I don't think I'm imagining this). Soundstage is a bit more coherent. This seriously challenges 730 now (may not be as punchy, but warmer) and even Tandberg TR-1040 (may not be as "holographic", but close enough). So a top-3 place in my modest vintage collection, which is much better than expected!
 
Wanted to report a few succesful tweaks, that I don't necessarily recommend, but resulted in yet another improvement for me.

Installed 470uf caps parallel to the output caps (I had UKL 50Vs in stock). Heftier bass without going boomy.

Since output capacitances of the new drivers and outputs were significantly higher than the original, I took off the 220pf (Miller) caps from the VAS-transistors. There's no risk to it as there are still more capacitance than originally, and testing shows there's no oscillation. I installed BC549s to predriver and VAS-roles, as they are suited to it better than the sc1845s (although different pin layout is a pain). The end result is more airy sound (really!).

Oh, I also added more low leakage caps, and a pair of films replacing one pair of 1uf elkos on the preamp, so these contributed some.

Now my 'golden' ears may lie to me, but I would not have written this if I didn't consider this to be a success (partly to my surprise!). Not to forget that synergy with your speakers and listening room are very important, so this may not apply to everyone.
 
Increasing the capacitance value of the output caps in a cap-coupled amp design will alter the filter effect… and therefore what frequencies get through or not. So, more capacitance in this design will result in more bass emphasis.
 
dlucy said:
Increasing the capacitance value of the output caps in a cap-coupled amp design will alter the filter effect… and therefore what frequencies get through or not. So, more capacitance in this design will result in more bass emphasis.
Click to expand...
Indeed. 1000uf in parallel was too much already, unless one prefers boominess...

Inspiration for these latest tweaks came after A/B-testing against 730. Noticed that the mids were about on par, but the low and top ends could do some work. Now it is pretty darn close, but I have to A/B-test them again.
 
JyriK said:
Indeed. 1000uf in parallel was too much already, unless one prefers boominess...

Inspiration for these latest tweaks came after A/B-testing against 730. Noticed that the mids were about on par, but the low and top ends could do some work. Now it is pretty darn close, but I have to A/B-test them again.
Click to expand...
You understand the 730 is a direct-coupled design and not really comparable to the 330B, right?
 
dlucy said:
You understand the 730 is a direct-coupled design and not really comparable to the 330B, right?
Click to expand...
What makes you think that sound qualities of different topologies cannot be compared to each other?

Personally I find it fascinating that this simpler design can challenge the heavies.
 
Last edited:
JyriK said:
What makes you think that sound qualities of different topologies cannot be compared to each other?

Personally I find it fascinating that this simpler design can challenge the heavies.
Click to expand...

I think it is fine to compare, just wanted to be sure you knew these two units had different.

Simpler often yields better results. Less "stuff" in the path to disturb the signal.
 
You must log in or register to reply here.
Back
Top Bottom