Krohn Hite 4400 (ultra low dist. sine osc.)

I bought a pair of Krohn Hite units for $100 total.
5800 function generator
4400 Ultra Low Distortion Oscillator
Neither had their fuses, and with new fuses, lights came on but both failed produce a signal.
Easy fix for both though.... the PSU electro caps and all tantalum caps were replaced.....and both came to life.
I bought the user/cal manual for the 5800. I cal'd it and it works well.

I cannot find the cal/schematic manual for the 4400.
This seems to be an early version, with 600 Ω outputs only. PC board date is 1980.....
The opamps employed are of the same era... glorified 741's and 300 series.....
MC1741 (x3)
LF13741 (x1)
LM310 (x1)
LM318 (x4)
....all directly soldered to the board, among a mixture of carbon and film resistors, etc.
Without a schematic or a Cal guide, I don't know how to adjust all the trimmers for lowest distortion.
I did pull the 741 types, and added sockets. Also, under the board I soldered some 100nf cer. caps to each power pin, to ground.
These opamps were lacking these, whereas the LM310's already have them.
Below, is the entire board, during the cap replacement, but before the 4 opamp swaps (on the left)

You might be able to see that i swapped to newer opamps, LF13741 AD845, MC1741 OP27. These were chosen from my spare chip collection.
There was an improvement, where the 2&3rd harmonics were now several dB lower.
The only way I can test distortion is using the FFT measurement on my R&S scope.
DC is removed via trimmers on the rear panel...obvious and easy.

I did not want to swap any of the 300-series opamps until I can find a proper manual and schematic. (I've looked everywhere).
That's where I stand. Anything you want to add is appreciated.
Last edited:
  • Like
Reactions: 1 user
I saw that earlier, but some links (Manuals Plus) produced the same dead-end results as my Goolge search , and another link looked sketchy.
However, just now, I carefully navigated through the " " link, avoiding all the spam-like ads that say "download", and I found the manual for the 4400A.
This manual seems to be the same manual also used for the 4400.
All the "A" suffix means, is there is an extra 50 Ω output. The rest of the unit is the same.
No wonder I could not find it.

Now, I'm working-out how to set-up the best way to measure distortion, and see if my KH 3202 Hi/Low filter can get me a steep notch with this adapter as suggested in the 3202 manual......
76MB and too big to print... here's a 9.6MB version....
Yup, same copy, more reasonable size. Thx for that.
I noticed a difference on my unit, to this manual's schematic.... My unit uses an LM318 for U141 , whereas the manual's schematic shows an NE5534a.
I'll socket that and fill it with a Signetics NE5534an....... and the 2pf comp cap seen in the schematic is already there.
Maybe I'll add the 50 Ω BNC output too.
Last edited:
Well it WAS working.......... then the Sine signal suddenly collapsed and disappeared from my scope.


Power supply is still working evenly with the new caps, power light still comes on.
(I took post-mortem DC voltage measurements at all the test points, and those are in the photo below.)
It's not dead, it's just in a coma.
Yes, I was messing with it..... solder-sucking the chips and inserting sockets and trying some NE5534a's...... but I was not touching it when it failed.

I replaced all the old chips in the new sockets, to no avail.
I was not sure if this old board was "through-hole plated", so I made sure to melt solder on both sides of the board at the socket-to-board pin junctions.
I also triple-checked this, checking continuity from all the opamp pins to the actual board traces. And I checked for pin-to-pin shorts.......nothing.
All opamps are getting correct power.
Nothing gets hot.
I went through the troubleshooting guide in the manual. It sorta points to "U153 and associated components".

I'm considering pulling the jFETS and testing them off-board.
But if you look at my notes on the image below...... the frequency buttons want to have either -15v, or +10v, depending on which button is it's missing +10v somewhere.....maybe one of the 10v zeners quit working ?

The scope is highly likely to have poor dynamic range compared to an ok soundcard. Some of the free spectrum analyser software has exceptional capability. That should allow tweaking and confirmation of harmonic levels in the 10Hz to 90kHz bandwidth, so not extending to your full equipment performance range, but very easy to setup and can be used with the notch.
Joined 2004
Paid Member
Attached is an improved scan of the schematic. Looking at your symptoms I suspect something around U153 is not connecting. The board will have plated through holes but removing parts could damage the board. PCB's have become much sturdier in the last 40 years. I would check continuity from component pin to component pin to spot open traces.


  • KH4400A schematic.pdf
    3.4 MB · Views: 23
  • Like
Reactions: 1 user
It's working again !!
I did not find any circuit breaks/faults, and I'm not exactly sure what fixed it.
All the zeners were good. I did replace about 4 jfets with the same NOS replacements.
I also put in 2 new trimmers, multi-turn type.
I may replace the rear panel DC trim pots with new ones.

While waiting for a few parts, I swapped-out most of the carbon resistors with metal film.
I think this 1979 unit was on drugs..... because early ALL of the old carbon resistors tested high, at almost 20% above what their color-bands stated.

After it began working again (all the old opamps installed), I did exchange most of the output/amp opamps with OPA1611 on adapters.
U125, I used AD8510.
U103 was LM310n voltage follower which has it's pin2 internally connected to it's output (pin6) I soldered a jumper connecting these two pins, so that with a regular opamp installed it's wired for unity gain. I assume that was the correct thing to do.

I'm waiting on some parts for a twin-t notch filter project...... so I can't really know any real distortion figures yet.
Last edited:
Joined 2004
Paid Member
This is the spec for the distortion. Put it in context of when this was designed-1980. Instruments with significantly lower distortion really did not exist. My modified KH4400 measures -115 dB THD+N @ 1KHz 10V with a 30 KHz low pass filter. THD measures -125 (really these are the limits of the Shibasoku 725D). The Victor oscillator is better, but single frequency and limited output voltage. Either is good enough to test audio products at levels far better than known hearing sensitivity.
KH4400 Distortion.JPG