Siglent SDM3065X Low Buzzer Volume Repair/Hack Attempt

[t1d]

Prologue
Note/Disclosure
- I am moving this topic out of the Siglent Bench DMMs SDM3065X, SDM3055 and SDM3045X thread, in the hopes that it will make the information easier to be found and to be followed.
- This first posting includes some duplicate information submitted in the prior thread. Included in the duplicate information, you will also find the links for information that I posted in prior years. You are welcome to review that, however, I intend to post complete details, in this thread.
- I know that others have worked on this matter. I acknowledge that I am building on their work and proclaim my deep appreciation for them graciously sharing their knowledge. I also know that I will not have found all of that information, meaning that I may find myself addressing subjects that they have already worked through and resolved. In such cases, please bear with me.

Here we go...
https://www.eevblog.com/forum/testgear/siglent-new-bench-dmm-sdm3055/msg5818373/#msg5818373
On Page 81 @ Post #2005
Re: Siglent SDM3065X - Firmware is current @ V3.01.01.12R1

Original Post
https://www.eevblog.com/forum/testgear/siglent-new-bench-dmm-sdm3055/msg4144669/#msg4144669
Beginning at Page 46 @ Post #1134
Ending at Page 47 @ Post #1152


I would like to (finally) address my units low buzzer volume. I have found a convenient mounting location, on the faceplate. I want to discuss 1) any consequences of using that location and, 2) if there are no problems with using the location, how best to wire the buzzer.

Location
I have discovered that there is a sixth meter probe hole, in the front, metal, face bracket. It is below the black, LoSense Probe Hole and to the left of the Red, 10A RMS Probe Hole. See picture. This hole is perfectly sized to accommodate the buzzer. I have not removed the face bracket, yet, but it appears that there may be space to tuck the buzzer in between the front of the metal, face bracket and the back of the plastic front panel.

Note: The hole in the metal, face bracket is covered by a very thin shim of black plastic. I confirmed the hole's location, by shining a flashlight through the plastic faceplate.


Accommodations
A hole would need to be opened in the plastic faceplate. For appearance sake, I recommend only making the opening the size of the small hole that is on the top of the buzzer. I don't think a larger hole in the faceplate will make the buzzer louder, because the hole in the top of the buzzer is the limiting factor.

Wiring
I am concerned with this hack introducing noise into the case. So, I am considering the following method of wiring...

Common copper solder wire wick is woven in the shape of a tube. I would use it to create a shielded wire loom in which to run a twisted pair of wire, for the buzzer. The wire loom would be grounded to the case/Earth GND. Care would be taken to not ground it to Circuit GND. I will cover the loom with heat shrink, as well.

There appears to be room to attach the loom to the side of the case. However, it will have to be routed under the fan housing. There is space to do this, as the PCB is cut out, under the fan.

Attachment
I would simply glue the buzzer to the metal, face bracket. I prefer a product called Shoe Goo.

I apologize for the lack of case pictures. Health problems prevent that effort, presently.

I look forward to your thoughts and suggestions. Thank you for your help.

===============================================================================================================
On Page 81 @ Post #2006
tautech shared these thoughts:
"Use a audio twin coax for the buzzer cable.
Maybe just gluing the buzzer to the front panel and nothing else will increase its volume."

@ tautech
I found a nice space, between the plastic faceplate and the metal, face bracket. I put these details in the post, above, so you will need to read that post, again.
Great ideas! Thank you!
t1d
===========================================================================================================
On Page 81 @ Post #2007
You might also consider a replacement/new buzzer, maybe one with a louder output.
Dunno anything about these things but other members will have some ideas.
The coax shield would be tied to Earth Ground, not Circuit Ground; correct?
============================================================================================================

End of Prologue

[t1d]

Here's a picture of the space behind the plastic face-plate, for the new buzzer location:

[KungFuJosh]

Has anybody found a part number for the original buzzer? The first thing I would do is to see what the dB rating is. If it's not that high, simply replacing it with something higher would be the easiest solution.

Have you tested the voltage for the buzzer? You'll need that info also.

If you have to move it to the front, something like this would probably squeeze in: https://www.mouser.com/ProductDetail/Same-Sky/CMI-1295-05TH?qs=olJun0bQHM9Fgf9TAxQbJg%3D%3D

[t1d]

Has anybody found a part number for the original buzzer? The first thing I would do is to see what the dB rating is. If it's not that high, simply replacing it with something higher would be the easiest solution.

Have you tested the voltage for the buzzer? You'll need that info also.

If you have to move it to the front, something like this would probably squeeze in: https://www.mouser.com/ProductDetail/Same-Sky/CMI-1295-05TH?qs=olJun0bQHM9Fgf9TAxQbJg%3D%3D

Thanks for helping, Josh!

1) There are no markings on the Buzzer = Nothing to identify its part number. Teaser = My buzzer is defective. The nature of the defect leads me think that they are all bad and deserve to be recalled; truly. I am presently writing it up. However, I am not young = health challenges = no sleep = folks are just going to have to wait for me to finish the text and make the pictures. I know somebody will complain, loudly, but that won't speed me up = grin.

2) The Buzzer has continuity on its positive pin to the +5V supply. Schematic and PCB pix to follow.

3) The measured dimensions of the buzzer are:
Diameter = 16.75mm
Height = 8.00mm
Pin Spacing = ~10.36mm
Pins are flat =~0.66x0.38mm
At 8mm high, the original buzzer case might barely be short enough to fit where intended. I need to do some more measuring. Your linked buzzer is too tall. I will be talking about just using a bare wafer; you don't need a case. You could also cut the inside of the face plate, but, personally, I do not want to do that. I would rather make things reversible to original, when possible, and this hack can accommodate that goal.

You also need to know what Hertz level that you can hear best. For me that is 400-600Hz. I have seen 400Hz wafers. It would be nice, if Siglent would code in a good many more frequencies. Isn't the lowest setting 100Hz; I don't recall the settings, right now. If so, IMO, that is too low to be very universally useful. So, maybe that could be changed to 400Hz.

Must simply make myself stop and rest, but I don't want to... This is way too much fun! Cheers!

More to come. Cheers!

[t1d]

Oh, yeah... The dB is not marked. I guess someone could hook up a function generator, mic and scope and sweep it. But, I am not up for that. More cheers... Told you I couldn't stop = grin...

[t1d]

Here is a picture of the schematic:

Here is a picture of the drive circuit, on the PCB:

[t1d]

Here are pictures of the buzzer:





Things to Note:
- It is passive, having no internal circuitry.
- Its positive brush pin terminates on the Piezo’s Ceramic Coating and its negative brush pin terminates on its Outer Ground Ring.
- The Ceramic Coating is significantly tarnished.

The stain is tarnish, not dirt; IPA/99% did not budge it. Tarnish = poor conduction = low power = low volume (and might slow code recognition?)

As I am a hobbyist and I have used the unit gently and only occasionally, it is clear to me that the tarnish is a manufacturing defect. This one cheap component (in what is meant to be a precision instrument) is likely to be the reason for our desperate low volume woes. We’ll see if Siglent owns up to the problem, issues a recall and supplies quality replacement buzzers. They are easily replaced. Hmm... I think it might be interesting to test the resistance of the tarnish... Oh, my unit was purchased before the Covid shortages.

Then again, there is also the matter of the buzzer’s location, to boot. I could try replacing the Buzzer, in its present location. But, as I have dug this far into the unit and my ear’s need all of the help they can get, I think I will stick with moving the buzzer to the front panel.

Completing the panel location installation would be less complicated, if the buzzer housing was shorter. As there is nothing in the housing, except for the Piezo wafer, I will just buy a bare wafer.

I already have a small wafer that would fit. However, I am concerned that its diameter is too small to create a tone low enough for me to hear it. I will just use my frequency generator to test some options.

The MCU/PWM driving method explains…
1) Why the continuity recognition time is so painfully slow,
2) Why the volume is at full throttle and
3) Why Siglent can only change the Buzzer’s frequency.

[KungFuJosh]

I am not young

🤔🤔🤔 You sure the buzzer isn't working? 😉

Your linked buzzer is too tall.

You mentioned wanting to use the front panel for the buzzer. I wasn't suggesting mounting that to the PCB, I was suggesting running wires from the board to the buzzer. The buzzer I linked is 12mm diameter, so it might fit in the front panel hole with little convincing (measure first, they have other diameters available). You can try it out and see how loud it is before drilling. Even if you do drill, you could probably get away with a 3mm hole center of the mounting hole.

Conveniently, the buzzer I linked is also 5V.

Thanks,
Josh

[t1d]

I am not young

🤔🤔🤔 You sure the buzzer isn't working? 😉

Your linked buzzer is too tall.

You mentioned wanting to use the front panel for the buzzer. I wasn't suggesting mounting that to the PCB, I was suggesting running wires from the board to the buzzer. The buzzer I linked is 12mm diameter, so it might fit in the front panel hole with little convincing (measure first, they have other diameters available). You can try it out and see how loud it is before drilling. Even if you do drill, you could probably get away with a 3mm hole center of the mounting hole.

Conveniently, the buzzer I linked is also 5V.

Thanks,
Josh

"Sure buzzer isn't working" = That's a legitimate question = big grin!

Buzzer too tall...
1) Yes, the height limit I was mentioning is in regard to the location in the front panel space...
2) Yes, you could cut away some of the plastic to seat the buzzer, if the buzzer has a case. However, there is nothing inside the buzzer case, except the ceramic wafer. So, you could just buy a wafer, without a case, and mount that, without any alterations.
* For #1 and #2, see Blue Arrow in picture.
3) Yes, I had thought of a small hole - just the size of the hole in the buzzer case - as well.
You hit the trifecta = Great job! Thanks.



I have other pictures from inside the DMM case. I will post them somewhere and post the link.

[KungFuJosh]

The MCU/PWM driving method explains…
1) Why the continuity recognition time is so painfully slow,
2) Why the volume is at full throttle and
3) Why Siglent can only change the Buzzer’s frequency.

I don't agree about the speed being painfully slow. Raise the resistance level to 1k on the continuity test and you will see it's significantly faster. My Keithley is faster, but it defaults to 1k test level. When the Siglent is also set to 1k, it's much closer.

Buzzer too tall...
Yes, the height limit I was mentioning is in regard to the location in the front panel space...

Do you mean the 12mm diameter? Tall isn't a relevant measurement here. Safety jacks usually have 12mm barrels, but the flat sides might be an issue. What is the distance you measured between the flats?

The part is listed as 9.5mm tall, most or all of which would be inserted in the front panel. If you have a depth gauge, you can measure that too.

They do have some with a smaller diameter, but knowing what the space between the two flat parts of the mounting hole would help determine that. Of course, a rotary too with a sanding attachment could make the flats for you.

[t1d]

I reply within the quotes...

The MCU/PWM driving method explains…
1) Why the continuity recognition time is so painfully slow,
2) Why the volume is at full throttle and
3) Why Siglent can only change the Buzzer’s frequency.

I don't agree about the speed being painfully slow. Raise the resistance level to 1k on the continuity test and you will see it's significantly faster. My Keithley is faster, but it defaults to 1k test level. When the Siglent is also set to 1k, it's much closer.
I forgot about the difference in the settings. It is a really good point. I need to remember to adjust my settings. Thanks!

Buzzer too tall...
Yes, the height limit I was mentioning is in regard to the location in the front panel space...

Do you mean the 12mm diameter?
No, I do not mean the 12mm diameter. A 12mm diameter might fit without any adjustment. The original buzzer diameter is 16.75mm. IIRC, it fit in the total available space, but it did not center over the probe socket hole. I can re-check that, but not today.

Tall isn't a relevant measurement here.
Yes, tall is a relevant measurement, because the height turns to the vertical plane and must fit between the plastic Probe Socket frame work of the front panel and the metal, face-plate of the DMM case. I hope that is clearer. If not, let me know and we will keep working at it... I can make a drawing, but not today...

Safety jacks usually have 12mm barrels, but the flat sides might be an issue. What is the diameter you measured between the flats?
Ahh... I get you, now... Height is not an issue, because you want to drop the case inside the flats... Okay, see below...

The part is listed as 9.5mm tall, most or all of which would be inserted in the front panel. If you have a depth gauge, you can measure that too.
No, I have not measured the probe socket depth. Remind me to do that, but not today = See the theme, here? grin.

The original case is only 8mm tall and it is really tight, between the plastic front panel and the front metal DMM case plate, as explained above. As a matter of fact, I did not try to close it up all together... I just eye-balled it. So, at this point, I am just guessing. But, as I said, there is no need to have a case on the buzzer, at all. Just use the existing wafer out of the original case, or buy a new wafer...

They do have some with a smaller diameter, but knowing what the space between the two flat parts of the mounting hole would help determine that. Of course, a rotary too with a sanding attachment could make the flats for you.
Probe Socket Front Face Hole = ~14.85mm
Socket Hole in Metal Front Case Plate = ~13.11mm
Flats - Inside Diameter = 13.25mm. Wall thickness = Less than 1mm, each. Then, there is a couple of more mm of distance, until you run into the PCB. So, a 12mm outside diameter case should drop in between the flats okay.  ANd, I think the probe socket barrel is deeper than 9.5mm. Remind me to measure it. But, why pay for a case that is going to make much more work, when a wafer is all that is needed? That's my thinking, anyway...

YouTube has videos on soldering wires the the wafer. It doesn't look too tough to do. Or, just buy a wafer that comes with the lead wires already attached.

Speaking of wire, I want to talk about what is needed, in terms of isolation and how to route them though the case. But, I will make that another post...
Cheers!

[KungFuJosh]

It sounds like the part I chose should fit almost perfectly. Maybe enough room for some double-sided tape to secure it in place in the front socket mounting hole.

Why pay for the case? To insulate the new buzzer, and make life easier, and the old case won't fit in that space. It's 99 cents including the case! Geez, how old are ya? 😉

[t1d]

How old am I? My first job didn't even pay $0.99/Hr... 
I will be steering clear of the Asian Online Markets, for this purchase, because of the very problems that I have discovered with our buzzers. Yep, OEM and, likely from Mouser.

Okay - The depth of the plastic probe socket hole, between the flats = ~10mm.

Things to consider about Piezo Ceramic Wafers...
Working voltage
Dimensions
The Frequency they produce - Center and total bandwidth.
The Frequency that you can hear.
Power - Active, or Passive. We want Passive.
Lead wires, or not.
If in a case, how is continuity made to the wafer. Soldered leads are much better than the brush pin type of our original OEM buzzers.
How you will mount just a wafer... You want to secure it, but not inhibit its ability to vibrate freely.
YouTube has videos on these considerations, and others that I have forgotten to mention...
Cheers.

[KungFuJosh]

The link I sent is from Mouser for the part for 99 cents. It fits in the front hole easily, and you can run wires back to the board. You've now established that the 12mm x 9.5mm case is ideal.

By using the unit that fits the space perfectly, and comes with its own cover, you don't need to worry about the rest except whether or not you can hear it. For 99 cents, connect the thing to a PSU and test it out.

[t1d]

The link I sent is from Mouser for the part for 99 cents. It fits in the front hole easily, and you can run wires back to the board. You've now established that the 12mm x 9.5mm case is ideal.

By using the unit that fits the space perfectly, and comes with its own cover, you don't need to worry about the rest except whether or not you can hear it. For 99 cents, connect the thing to a PSU and test it out.

There no fallacy in that thinking. I will add some to my pending Mouser Cart.

DS says the wafer uses 30mA. Vsup may vary from 3 to 7 volts. Vsup effects its frequency output, too = 2100Hz to 2700Hz...
Resistor values are both 810R, but, both of my resistors measured 1K, in circuit.
Do you have the skill set to calculate, or simulate, the existing circuit, in order to know what actual voltage and current are available to the wafer? I need the numbers, to be able to set up my function generator, to do some testing. I'm too fogged to really do the math, today...
Let's see... 5Vsup - 1.4V Transistor Drop = 3.6Vactual
Yep, totally fogged...

MCU AM3352BZCZ60 = DS Link = https://www.ti.com/lit/ds/symlink/am3352.pdf?ts=1739828310159&ref_url=https%253A%252F%252Fwww.mouser.com%252F


Cheers.

[t1d]

I believe the transistor is a BC846B. DS attached.

[KungFuJosh]

Sorry, I can't, and I wouldn't guess at the circuit either. There are certainly members here who wouldn't have much trouble figuring it out.

[t1d]

Here is some wire that I have = https://www.amazon.com/dp/B0D2QHXMPY?ref=ppx_yo2ov_dt_b_fed_asin_title&th=1
Spec sheet attached.
I think it will do nicely.

I need instructions on how to minimize adding noise to the inside of the DMM case... Dos and Don'ts... Routing along the case wall and around the fan, or under the PCBs, or fly it over the PCBs? Use individual cables for positive and negative? Ground the Buzzer signal to circuit ground, yes? Ground the coax shield to case/Earth ground (only) yes? I have never had to deal with noise at this level of importance...

[KungFuJosh]

I would assume the buzzer has 2 connections, which you should reuse. I wouldn't introduce new connections for that at all.

If it's recommended that you use a coax shield, I assume that would go to chassis ground, and shouldn't interfere with the circuit at all.

[t1d]

I would assume the buzzer has 2 connections, which you should reuse. I wouldn't introduce new connections for that at all.

On this point - 2 or 3 connections/wires - you need to be careful. IIUC = No guarantee = Two connections/wires are for passive and active piezo wafers = That is what we want.

Our DMMs are a little tricky... They have switching circuitry, nearby, but no capacitor/inductor relaxation oscillator components.

That might lead you to think that the active oscillation components are in the buzzer case (as they are with an active buzzer.) Because I 1) opened the case, to verify that there are not any vibration components and 2) followed the traces back, in the direction of the MCU, we have a better guess that the MCU is providing the oscillating signal that drives the wafer.

The MCU signal could be proven with power-on testing and a signal analyzer/oscilloscope. I find it unnecessary to do that test, because we are just replacing the wafer with the same type.

Three connections/wires is for self-driving/vibrating buzzers. IIUC, they do better at centering the resonant frequency, to get the loudest volume. They require their own research.

Ceramic piezo buzzer wafers have their own learning curve. Understanding and attention to detail are required, as with all electronic components.

If it's recommended that you use a coax shield, I assume that would go to chassis ground, and shouldn't interfere with the circuit at all.

Yep, that's my thinking. But, the details of routing it properly are extremely important, too. Hopefully, someone knowledgeable about those details will come along and help us.

[t1d]

I am keeping my thinking open, about solutions. To a great degree, the proper solution is rather personal to the owner of the DMM. What skills and resources do they have... How well do they hear... Etc. Aging has changed all of those factors for me.

A resource that I have in hand is a Piezo wafer that is already in a very flat shell. It is a Kyocera type. I think it was meant for printers/copiers. It looks very much like the microphone cartridges in the handsets of old dial face telephones. (Yes, Josh, I really am that old = grin.)

In the vein of creating an easy solution, I am considering this approach... The original site for the buzzer is near the rear of the case. There is a safety keyhole cut in the rear case plate, which I will never use. I could:
1) Solder pin sockets in the PCB's buzzer pin holes. (More about this to follow.)
2) Run the wafer wires through the safety keyhole.
3) Mount the wafer to the back of the case with double-sided tape. The wafer is so flat that I believe this will appear to be an OEM/original installation.

Using sockets for the buzzer pin holes allows me to test the rear case buzzer location. Should I still need to move the buzzer to the front of the case, removing the rear buzzer connection would be easy, as would plugging in the new wires to go forward. It is worth a shot...

[KungFuJosh]

I grew up with rotary phones too...back when the phone company rented them to you, before you could easily get your own. But I'm not thaaaaaaaaaaaat old. 😉

The piezo on the rear idea is interesting. Try it out; if it sucks, undo it. Easy!

[tautech]

Sorry, I can't, and I wouldn't guess at the circuit either. There are certainly members here who wouldn't have much trouble figuring it out.

Some little Gurgling about piezo buzzers reveals a lot.
Normally driven with a PWM signal of just a few volts which might cause some interference to sensitive circuits which would be good reason Siglent placed it on the digital PCB instead of the analogue PCB.

Still, digital signals go to the front panel via just a ribbon cable carefully placed against the internal chassis wall.
A shielded twin coax to the buzzer should work just fine.

[t1d]

I threw this MuRata PKLCS1212E2400-R1 into my Mouser cart, just to have some additional options.
https://www.mouser.com/ProductDetail/81-PKLCS1212E2400-R1

From the Data Sheet:
https://www.mouser.com/datasheet/2/281/PKLCS_datasheet-1186506.pdf
Warning = No sleep, again = Brain Fogged thinking... And, a whole lot of frustration... Argh... grin.
1) Per #3, below, do I need to be worried about DC voltage? I have talked myself in circles = The MCU sends a PWM signal that swings from 0 volts to 5 volts. So, is that a DC problem, or is it an acceptable AC signal that just happens to be offset to only swing on the positive side of the AC cycle. Sorry for the noob confusion...
2) Per #4, below, given that our resistor is 810R (but measures 1K, in circuit,) if I decide to use this buzzer, should I change/increase the resistor value? If so, what value would be best, between 1K and 2K, and why?

Yes, Josh, I grabbed three of the type you suggested = thanks for telling us/me about it.

FYI = Mouser has a pop up stating that there may be some weather delays in shipping.

[t1d]

I fooled around with some of the piezos that I have... Sort of comparing them by ear as to volume (which is really pretty useless,) given the same frequency and voltage = 1KHz @ 1 volt.

Compared to the original buzzer, the large Kyocera may be slightly quieter. It produced some of the lower harmonics better. Maybe the lower tones ate some of the current? It is a nicer tone, though.

I have a tiny wafer, too... Maybe 8mm diameter. I couldn't get it to sound, until I ran the frequency up to 8-10KHz. Yeah, that's useless. I was surprised that I could hear it. I wonder what it is actually intended for? Maybe smoke alarms, or pest/mosquito deterrent gizmos.

I intend to do some more serious testing, when I feel better. I am looking forward to that.