RetroBrew Computers Forum: General Discussion

Home » RBC Forums » General Discussion » faint speaker (very low volume on SBC V4-004 circuit)

Show: Today's Messages :: Show Polls :: Message Navigator

faint speaker [message #8690]

Fri, 18 June 2021 11:20

lynchaj
Messages: 1080
Registered: June 2016

Senior Member

Hi
I am having a time debugging the speaker on the Z80 clock board. It is the same circuit as on the SBC V2-004 board which is in turn based on the TRS-80 beeper circuit

https://www.retrobrewcomputers.org/lib/plugins/ckgedit/fcked itor/userfiles/image/boards/sbc/sbc_v2/sbc_v2-004/sbc_v2-004 -snd.png

On the first Z80 clock board I made the speaker sounds just fine. Not to loud, not to faint, it worked pretty well. I was really happy with it.

However on the second Z80 clock board the audio volume is extremely low. I can hear it if I listen closely but because its so faint it is not useful as an indicator.

I've replaced every part of the circuit starting with the speaker, the 2N3906 PNP transistor, etc. I've measured and re-measured the resistors and they're all the right values and connected properly. Nothing seems to increase the speaker volume. Except when I put my oscilloscope lead on the emitter of the PNP it seems to get a bit louder. Not much but noticeably louder.

This is the piezo speaker I am using:

https://www.jameco.com/z/EAT-17-Projects-Unlimited-Speaker-P iezo-3800Hz-80dB-20-Voltp-p-1mA-2-Wire-6-9mm-x-14mm_2151216. html

I did some experimenting by carefully lowering the values of the resistors (less than 50%) to improve current flow and no effect.

Also I wrote a program called tone3 which I can run from CP/M to play a tone for about 30 seconds. It is approximately 1500 Hz. I know it is running because I can hear it start and finish and also it turns on the user LEDs just before it finishes to indicate it is done. It finishes normally.

Any ideas as to why this the volume is so low on this board but on an almost identical board it is much louder?

Thanks, Andrew Lynch

Report message to a moderator

Re: faint speaker [message #8691 is a reply to message #8690]

Sat, 19 June 2021 04:57

b1ackmai1er
Messages: 396
Registered: November 2017

Senior Member

Hi Andrew,

Installed back to front?

Kicad foot print issue? Try changing the foot print to Package_TO_SOT_THT:TO-92_HandSolder (as per v004) temporarily and see if it does anything unexpected with the routing/ratnest.

Also the Nascom rom basic can play tones - PLAY Octave, Note, Duration (0-8,0-11,1-8) (not my best work lol)

Edit : Piezo installed with correct orientation ?

Edit #2 : After rereading your post, I realize above advice not really applicable since you are comparing two identical boards, not MCB vs v004 - Sorry, best of luck.

Regards Phil

[Updated on: Sat, 19 June 2021 05:08]

Report message to a moderator

Re: faint speaker [message #8693 is a reply to message #8691]

Sat, 19 June 2021 06:24

lynchaj
Messages: 1080
Registered: June 2016

Senior Member

Hi
I have tried flipping the orientation of the piezo buzzer/speaker and that doesn't seem to have any effect. I think it is already reversed with the + pin grounded and the other pin getting the signal. On the first Z80 clock board it didn't make any difference and it works just fine. However this afternoon I'll try flipping the transistor in case there is a subtle footprint error. However, it is pretty clear from the schematic the emitter is VCC and the collector connects to the 120 ohm resistor. It is possible I have these flipped somehow so I'll give it a try and see if things improve.

I am also wondering if I purchased 2 bad piezo speakers because I swapped them and got the same results. This happened with the 74ls07 problem on the Z80 processor board. Swapping parts doesn't necessarily mean getting a good part if its from the same batch. I think I need a dissimilar speaker and try that too.

Thanks, Andrew Lynch

Report message to a moderator

Re: faint speaker [message #8696 is a reply to message #8693]

Sat, 19 June 2021 17:09

lynchaj
Messages: 1080
Registered: June 2016

Senior Member

Hi

Well some experimentation today produced some good results. I tried flipping the transistor in case I had some how misread it. However placed in backwards the transistor blocked everything and I didn't even get the tiny volume I had before so I switched it back to the other way. No harm done and it rules out that possibility.

I've been comparing the first Z80 clock board with the second. One thing I found was on the first board when the tone program is playing there is a pretty good voltage drop across the terminals of the speaker. Like about 4 volts or so. It has a nice tone volume and works great. However on the second board with identical parts I measure just a tiny fraction of a volt drop across the speaker terminals. Like way less than 1 volt.

The speakers are identical parts and all the components behind it are the same as well. So what the difference? I don't know but decided to try putting a 10K resistor across the speaker terminals and the volume suddenly picked up to a normal level. I had noticed that putting the oscilloscope across the speaker terminals improved volume a tiny bit but the 10K resistor really made it sound great. I am starting to suspect the speakers in the second batch are not so identical to the first one I bought.

So at least there is a fix. I am going to solder in the resistor tomorrow and call it good enough. Place a note on the schematic that if speaker volume too low to add a resistor across the speaker terminals to boost it up a bit.

Thanks

Report message to a moderator

Re: faint speaker [message #8697 is a reply to message #8696]

Sun, 20 June 2021 16:40

tingo
Messages: 115
Registered: February 2017
Location: Oslo, Norway

Senior Member

Different impedance on the speakers perhaps? You should measure the "good" and the "bad" to find out the difference. Just use your regular ohm-meter (on your multimeter).
Still, it seems unusual. Home sound system speakers are normally low impedance (4 - 16 ohms), but I see that the image of the schematic in the first post specifies a 120 ohm speaker.

The specifications of the piezo buzzer you linked to doesn't say what impedance it has, but if we say 20 volts and 1 mA, it is about 20 kohm (simplified, just using ohms law). So maybe the 10 kohm resistor helps to load the transistor enough.

Torfinn

Report message to a moderator

Re: faint speaker [message #8698 is a reply to message #8697]

Mon, 21 June 2021 07:25

lynchaj
Messages: 1080
Registered: June 2016

Senior Member

Hi
That's what I suspect is the speaker for the first board is different somehow from the second board. Why, I don't know but it wouldn't be the first time a "standard" part varies from batch to batch. I did try measuring the impedance of the speaker (it's piezo) and with the regular ohm meter it seems to be infinite so it may be an impedance type thing.

The PNP provides some simple amplification so you're probably right about loading down the transistor. Probably doesn't have enough drive to push this speaker but by putting the 10K resistor in parallel it allows more current flow out the emitter.

Regardless, it works now and I put a note in the schematic for future builders. It is a mystery to me though. Strange!

Thanks, Andrew Lynch

Report message to a moderator

Previous Topic:	Working Text Editor for SBC188
Next Topic:	CPM HDDs on SD Cards - Implementation

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Tue Sep 30 11:55:17 PDT 2025

Total time taken to generate the page: 0.00773 seconds