Hohner Bass 3 (not really a pedal)

[FUZZZZzzzz]

A friend of mine has been looking for one of these keyboards / synths if you like. It has three sounds (guitar/tuba/strings) and some sort of decay and is completely monophonic. But, they sound pretty cool (to me). Sometimes they pop up on eBay, but not enough to keep the price low.

Yesterday I happened to find the schematic by accident and was amazed by how simple it really is.

I'm going to try and build this without the keys. I want to etch a sort of stylophone keys layout using the different resistors to tune the separate keys. Maybe even trimpots and keep this as small as possible. I would like to be able to power this with a 9 v battery as well. Theres a website where someone wrote that there was an error in the drawing of the power section. Dont know if someone can spot it (if true).

The sounds should work. Theres someone on youtube that breadboarded the different sounds. Just need to update the BC148b with something more available.

Hope its ok to post this here. Its not really a stompbox, but it has a guitar preset

[FUZZZZzzzz]

schematic link

https://i.postimg.cc/D0d85N7F/Hohner-Bass3.jpg

[Slowpoke101]

Interesting little circuit - could be fun to play with.
The power supply error with the schematic is that pins 2 and 3 of the 78L05 should be transposed (swapped ).
Since the entire circuit operates at +5V (ish - the pitch pot would vary this slightly ), you should have no problem with powering it from 9V. Check the current required though - a 9V battery may drain pretty quickly.
The BC148b transistors can probably be replaced with almost any of the usual NPN BJTs that we use (2N5088, 2N3904, BC548, etc ).
Have fun with it.

[anotherjim]

...Just pop the direct link in the forum image tags.

I didn't know about this thing. Cool key range on the original of E-E putting a whole octave in the middle - much more useful for bass parts.

If you can't get a 74C14, a CD40106 could be more like it than a modern 74HC14...


Those weird resistor values in the keyboard could be a bother. The old 2% metal film resistors used to come in odd values like that.

[FUZZZZzzzz]

thank you guys..  my supplier (sounds a bit ackward) stocks the 40106. that was the only thing i couldnt find. the resistors values are quite weird indeed. i think a couple of trimpots would be easier to tune.  Also, it is supposed to be amplified through a guitar amp (just like the hohner pianet) ... I might want to try and include a preamp if the signal is too low.

I looks like the sound presets only differ in cap values. could be cool to add an extra sound.

[Slowpoke101]

Ah...I've made a mistake (what else is new? )
Upon further examination (thanks to anotherjim ) of the schematic, I can't really see any error other than the 78L05 pinout diagram is different from what I've seen on the datasheets. Pin 2 is normally the centre pin but in this diagram it is shown as pin 3. As shown, the power supply should work fine.
Oh well.

[FUZZZZzzzz]

I was wondering about the fact that it has both a tuning and a pitch knob. pitch is probably for a wide range and tuning more of a finetuning?

[Rob Strand]

thank you guys..  my supplier (sounds a bit ackward) stocks the 40106. that was the only thing i couldnt find. the resistors values are quite weird indeed. i think a couple of trimpots would be easier to tune.

I think the idea is you use precision resistors so you only need a single global tuner.

The resistor values are E48 series or above (as opposed to the common E12 series and E24 metal films),
http://logwell.com/tech/components/resistor_values.html

When you buy precision resistors they often come with those funny values, especially if you buy from larger suppliers.

[FUZZZZzzzz]

What do the snowflakes next to r1 r2 and r3 mean? I have never seen this symbols before on a schematic.

[Slowpoke101]

The "snowflake" symbol is probably a "M" on its side. The M representing a metal film resistor. High precision resistors are usually metal film types.
A CMOS 40106 IC would be suitable as a 74C14 - almost the same thing if not actually the same.
The pitch control would make a note sharp or flat but the tuning control would have a greater range.

[Rob Strand]

hat do the snowflakes next to r1 r2 and r3 mean? I have never seen this symbols before on a schematic.

Normally when you see those there will be some text (somewhere) on the same page with a "*".  After that star will be some text.    Unfortunately I can't see such text.    It would be like,

* See calibration procedure.     ;The calibration procedure is written on another page.

or

* Adjust in the order RV1 ... RV ...

Or, the text following the * is the calibration procedure.

[Rob Strand]

he "snowflake" symbol is probably a "M" on its side. The M representing a metal film resistor.

It could even mean 10 turn pots.   (all speculation)

[Slowpoke101]

I think Rob was correct with the "*" symbols meaning a calibration or set-up adjustment.
Without a service manual or adjustment notes you would just 'wing it' and adjust for what seems to work with minimal smoke - preferably with no smoke at all.

[Rob Strand]

Go here,
https://archive.org/details/sm_Hohner_Bass_3_Usermanual_and_Schematics

The left hand side of the last page under "GENERATOR PLATE" it shows the text with the "*" 's next to it.

From what I can see it's just some descriptive text.

[duck_arse]

extra points for finding the BC148's, in their X09 lockfit package. seems they are directly related to the BC108, and became BC548's in the TO-92. [it's amazing what you can do w/ a couple of phase shift oscs.]



my guess on the power supply error is the single pole switch on the mains.

[PRR]

> the resistors values are quite weird indeed.

They want to be log-scale, each one twelfth-root of two higher. Same as your guitar frets get closer as you go up in pitch. (In fact if you could find a resistor as long as your guitar string active length, and cut it at the frets, those would be the right values.)

The magic number is 1.0594630943592952645618252949463

This to give equal *pitch* intervals with several good approximations to perfect intervals.

Using 1% resistors you will annoy piano tuners and the better musical ears. Errors will run about 10 cents sharp/flat. 0.1% resistors are available at 10X the price. But buying several octaves of single values is costly.

FWIW: the alternative is to build an Exponential oscillator. (Usually a linear osc with an exponential converter in front.) Now all the resistors are the same. But the Exp function is fussy and drifty at pitch-precision. (Read any early MOOG/ARP story-telling- they spent half their time tuning.)

[Rob Strand]

Using 1% resistors you will annoy piano tuners and the better musical ears. Errors will run about 10 cents sharp/flat. 0.1% resistors are available at 10X the price. But buying several octaves of single values is costly.

For one-off jobs it's possible to measure E12 resistors and calculate a tweak resistor.  If the measured value is under what you want you add a small series series.  If the value is over you add a large parallel resistor.   Just put the values in a spreadsheet.    You can also chose not to add the resistors if the error is below a certain level.

Unfortunately a lot of cheap DMMs only have around 0.5% accuracy on resistance.   If you have a 4 1/2 digit multimeter the meter error is removed by forcing all measurements to be done on the same scale (say 20k ohm scale). For low value resistances you need you subtract the the meter's zero resistance reading from the measurements.

If you are going to place an order for parts anyway the 0.1%'s are worth saving the headache.  They aren't *that* expensive these days.

At the end of the day I'm pretty sure the manufacturer would have used low tolerance resistors.

The final pitch accuracy not only depends on the resistor tolerance but also on the initial error.  The standard resistor values aren't stepped at 0.1% steps.   So standard resistor values will be off from the start to some degree.    So technically you could come up with better values ... but that's involves redesigning the wheel.

[Rob Strand]

FWIW, I calculated the frequencies based on the resistor values and the frequencies were way off (like 20%).  I keep thinking I've got a bug but I can't see it.

Anyone else want to give it a shot?

The other thing I noticed is the lowest note is largely unaffected by the R5 tuner trimpot (500hm) so that means the low frequencies are tuned with the PITCH trim pot.  I have doubts there's enough adjustment there to handle 74C14 part variations so you might have to tweak the cap value to get that right.
-------------------------
EDIT:

I think all the keyboard resistors are reversed.  So the first resistor to the 33.2k should be the 2k0 and the one at the end (which goes through all the resistors) is the 10k2.

EDIT:
I can confirm the error.  When I use the corrected resistor arrangement the maximum error is about 0.18%.  On average the error is 0.07%. Using typical Fairchild 74C14 specs at 5V I need a 17.6nF cap for the correct tuning.  So the cap isn't far off the 18nF on the schematic.  (I think I've got the octave right.  The lowest E from the div by 4 output is 42Hz.)

[Rob Strand]

It occurred to me I can attach a spreadsheet to the post.   It's not much use to me.   
Save the file and change the extension from .txt to .xls. If you want an open office .ods file I can do that too.

You can see by some judicious tinkering the frequency error can be halved.
If you want to see the frequency errors of the original values just copy the "Rk original"
values on the right hand side into the "Rk" column.   To put the tuned values back again
just copy the "Rk mod" column back into the "Rk column".

The blue cells are the circuit values you can tune.   There's some values for these on
the right.  They are the best values which corresponds to the original and modded Rk values.

The L2 and L4 values in orange give an indication of the overall goodness.

[anotherjim]

To be fair, I think this is the kind of thing you could front-end with a cheap kiddies keyboard. Pick the cleanest tone and square it up to drive the dividers. The voltage regulator can be fixed for +5v if you do that since it doesn't need to tune it. It looks like it has a crude envelope detector (worked in with the percussion effect) which will help.

Other substitution - although the 4024 is easy, if you don't have it, any dual D-type or J-K flipflop can do the same job. A 74HC part can work IF the Tune control doesn't go above +7v.