In developing my own precision tuner and being interested in strobe tuners, I found a working-but-needs-refurbishment Peterson Model 400 inexpensively on ebay, so I bought it with the intent to fix it up, calibrate it, and use it. I replaced a few parts, made a couple modifications, and added an internal mic to make it quicker to setup and use.

How a strobe tuner works:

A strobe tuner, at least traditional ones, is an electromechanical tuner that relies on the strobescopic effect (surprise surprise!) for its display. What that means is that a strobe tuner converts an audio input into a blinking light - neon lamps in this one - and shines it on a wheel with patterned markings that a motor is spinning at a constant speed. When the rotation speed of the disk, which has its speed precisely governed by a separate circuit, and the speed of the flashing of the light match up in the right way, it looks like the markings on the disk are moving very slowly, or not at all. This is the same sort of effect you can see watching a video of a driving car's wheels, where the wheel seems to rotate quite slowly, but in this case, the designers have done all the math needed for the speed of the motor and the patterns on the disk to look like they stop moving when the incoming audio frequency is in tune. It sounds complicated, and I suppose it is, but the circuitry involved is comparatively much simpler than other electronic tuners, so when electronics were comparatively primitive, it could be an extremely accurate instrument without too much electronic complexity.

You set the note you want to play, make sure the vernier knob is set to the right reference pitch, and then play your note - when close to the right pitch, you will see some of the markings on the disk get darker. If the markings are moving to the left, your note is flat, if they are moving right, your note is sharp, and if they appear but are not moving at all, you are perfectly in tune. The rings on the disk match up to different sounding octaves, so the higher the pitch being measured, the higher up on the disk the pattern will seem to slow or stop moving, and because the tuner measures all of the incoming audio, you will typically see multiple rings looking active - higher harmonics in the sound will show, usually more faintly, on the other rings of the tuner disk. Because the motor speed has to be precisely controlled to measure against a fixed reference, you have to change the note dial (centermost large dial) when tuning other notes, but because the harmonics show, you can still get good tuning information on the perfect fourth and fifth of whatever note the tuner is set to - the patterns will still be visible and will stop moving when the note is in tune. Interestingly, when playing the major third below the note the selector is set to, you will see one ring telling you that you're playing flat, even if you are playing the correct pitch. This is an artifact of equal temperament and the harmonic series, so if you play your note 12 cents sharp, that ring will stop moving, because you will be fitting in the just intonation of the major third. Conversely, if you set the vernier to 12 cents low, when you play your not a major third below and see the ring stop moving, you'll be playing the note in tune with equal temperament. The reference pitch vernier is marked in cents, but you can use it to play with a non A=440 reference pitch, for example, if you set it to be just over 7 cents high, you will be effectively playing with an A=442 reference pitch.

It helps in understanding the design and troubleshooting or refurbishing to have the schematic available, and it is available online, including here. The schematic even shows which board each part is on (dotted outlines), which is especially useful because there is no silkscreen or part markings on the boards, and it includes some of their design considerations - using low current noise resistors on the audio input line and using high stability, low leakage capacitors on a few key components used in keeping tuning frequency stable. To control motor speed, the tuner uses a high stability oscillator (at least, for the time) that drives a bipolar stepper motor, and each note on the selector has an individual resistor to trim the speed exactly. The audio processing goes through an optional low pass filter (the image clarifier switch) which I like to leave on, and then to the lamp driver board - to get that relatively small signal to drive neon lamps which have a 400V power supply. Later model 400s included an internal mic, but this one and this schematic do not include one, the input connector also does not supply any sort of bias voltage or phantom power.

Refurbishing the tuner:

If you're planning on trying this yourself, if you have some experience working with electronics and mains voltage electrical appliances, it's a fairly straightforward job - but even if you are, please be careful. Not only is there exposed mains wiring inside, there's also a high voltage 400V rail for the neon drivers and there's spinning parts thanks to the motor and disk, so there are a number of bits to be careful of. My unit's capacitors discharged fairly quickly when unplugged, but be sure to check them with a multimeter before working inside, even when powered down and unplugged.

I powered the unit up using the included mic and while it lit and whirred, it was entirely unresponsive. I used another external recorder's line out to run into it and suddenly the tuner sprung to life... but it was something like a fourth off of what the note was actually sounding. I knew it needed some help, so I went to work. Checking out the obviously dead mic I took it apart. I tried to get the element back on the little bar in the middle of the copper cup housing it, but since the mic wasn't dynamic, I'm not sure it ever could have been used with the tuner alone as no bias voltage was being supplied. I couldn't find any information on the mic model and wasn't expecting much, so I ended up tossing it.

Taking it apart, there are four screws in the wooden road case style housing and then the entire tuner chassis slides out. There are boards everywhere (the top is the power supply, the internal board drives the motor, the board on the side processes the audio, and the board that drives the lamps is tucked right next to the disk up front) and wires that fly all around, but because they are divide on the schematic and color coded, it's not too hard to find your way around. On the main board under some masking tape was a cluster of capacitors, and these turned out to be the tuning cap and one of the bunch had one leg not attached - probably broken because of vibration at some point. Reattaching it instantly fixed the problem of the tuner being very far off from the right note, now it was only a little bit off on some notes, which I think should be expected given the aging of components over more than 30 years - some part date codes suggest a manufacture date for this tuner around 1974. The first fix is an easy one!

I decided that the tuner would be much more useful with an internal mic, so I found a spot on the front place where you could add it, looked through the parts bin, and put together a little circuit. A switch connects the mic to the input line (if it were located differently, you could switch in the input or mic, but that would have required more cable and everything going right by the high voltage lamp driver board), then a small electret microphone is biased by a couple of thin film resistors and is powered off the 20V rail from the audio board. A couple holes in the front plate and a little bit of assembly later, it fully works and shows up well on the tuner's screen. Not as nice looking as some options, but plenty effective. I used a shielded cable for the signal and ground and the original design uses one to the input jack, but when I went to attach it to the jack, I discovered that their shielded cable used a foil shield which wasn't soldered to the jack, but just wrapped around. The jack was looking dirty and the cable going to the board needed to be replaced, so I just swapped all of it.

At this point, I thought I was done, but in giving the motor a second shot of contact cleaner, I noticed a slight tingle touching the front panel of the unit. I brought it upstairs (was in the basement) and tested again with no tingle, but hooked the multimeter between the chassis of the tuner and an earth terminal on a function gen, the chassis was at 60VAC. That potential was very, very low current, one microamp on the same meter (which is also one least significant digit), but it was enough to be felt, so I wanted to fix it. Reading around, I checked the capacitors on the primary and secondary sides of the transformer and while they measured well, one of the large filter caps on the secondary side had electrolyte bubble up from one terminal when testing it with the multimeter... so it needed to be replaced. I ordered up a proper three wire earthed plug and two replacement filter caps. Earthing the chassis was simple because there were already a few options as tie points, and replacing the caps meant trying to figure out how to hold them in place, but was quite easy to wire up. That eliminated the voltage on the chassis and prevented the incoming failure of those filter caps.

Calibrating the tuner:

I actually did the calibration before finding the 60VAC issue, but only because I had thought I was nearly done before I felt that tingle! Calibration of the unit is done with 12 potentiometers on the motor control board, one for each semitone of the scale, and most of them were slightly out because of the age of the thing. Procedure for calibration is fairly simple, but you have to have the wooden case off while it's powered up, so you do have to be a bit careful.

First, you feed a reference tone into the input. A simple sine wave will do, but any completely stable, completely accurate signal will work, and work out the math for the frequency of each with this formula: frequency = reference tone * 2 ^ (half steps from A / 12). The reference the machine states is A = 440 Hz, and the number of half steps can be positive or negative, the resulting number will be different, but the pitch will be the same, just in different octaves.

Then, locate the adjustment potentiometer for the corresponding tone and adjust it with a screwdriver (or better yet, a nonconductive trimmer screwdriver) until the darkest pattern on the display stops moving left or right. I like to let it run for a few seconds to see if there is very slight movement, but it will take some trial and error to get the adjustment right, the potentiometers are fairly coarse adjustments.

Repeat this for all 11 other semitones.

Since the board doesn't mark which adjustment corresponds to which note, I wrote in sharpie for each and included photos. It's not difficult to determine in testing, but it's handy to know in advance. It's also worth being absolutely sure that your pitch vernier is centered exactly at 440 Hz for calibration, as you can see in the earlier picture, my first full calibration was done 7 cents off until I realized it wasn't set right.

Now I've got myself a working, calibrated, vintage strobe tuner. The original spec list it as accurate to 1/3 of a cent, which is about as a fine as a person can hear and a bit better than my normal electronic tuner. There's a little bit of noise from the motor (which matches the pitch of the note you're set to!) and it's large for a tuner, but it's a well made bit of gear and a useful tool all the same. The visualization of a strobe tuner is definitely an improvement vs. your standard needle tuner to me.

December 6, 2016