Rohde & Schwarz SML Series Signal Generators - SML01 SML02 SML03 SMV03

[KungFuJosh]

I designed the conversion PCB for the MSWA2-50+ to SSW-224. PCBWAY can probably make them, JLCPCB can't (too small for castellated holes - they require minimum 10mmx10mm).

Somebody else please verify.

If you order these, I recommend ENIG plating, and make sure to use at least 25u for hole copper thickness.

And of course:
USE AT YOUR OWN RISK.

Thanks,
Josh

ETA: lol, it just occurred to me how tiny this is. Should be interesting to work with. 

[KungFuJosh]

eBay seller refuses to help me at all it seems, so I am kind of out in the weeds on this one hence why I started my own debug. He's been MIA for the past while. I guess I should have expected this though.

I know it's annoying, but given the price you paid for an SML03 that works fine except for the B3 option, I wouldn't worry about it. The PCB I made above should work to replace your bad switches.

Thanks,
Josh

[TylerPeppy]

eBay seller refuses to help me at all it seems, so I am kind of out in the weeds on this one hence why I started my own debug. He's been MIA for the past while. I guess I should have expected this though.

I know it's annoying, but given the price you paid for an SML03 that works fine except for the B3 option, I wouldn't worry about it. The PCB I made above should work to replace your bad switches.

Thanks,
Josh

Nice boards, thanks! I was working on some adapters in Altium yesterday as well, it was just taking a while in between gaps of work. While I shouldn't really be concerned, and can bypass the problem, I still need to get some rigid coax and solder-cup SMA connectors to put the fix together.

I am still working on figuring out the switch removal. I am having the issue where the thermal pad on the board is sinking too much heat into my table when desoldering. The white switches are more problematic than the black ones as the difference is the ground pour under the chips.

- Tyler

[KungFuJosh]

Nice boards, thanks! I was working on some adapters in Altium yesterday as well, it was just taking a while in between gaps of work. While I shouldn't really be concerned, and can bypass the problem, I still need to get some rigid coax and solder-cup SMA connectors to put the fix together.

I am still working on figuring out the switch removal. I am having the issue where the thermal pad on the board is sinking too much heat into my table when desoldering. The white switches are more problematic than the black ones as the difference is the ground pour under the chips.

You're welcome. Hopefully it helps somebody.

Removing those switches is probably easier on a hot plate, but you gotta be careful not to remove anything else. If I were to try it, I would use a hot plate with the chip quik kit I linked before. But, I'd probably fuck it up, so take that with a grain of salt.

[rf-messkopf]

Somebody else please verify.

I might be getting this wrong in my haste, but according to the schematic you posted you may need to swap the Control1 and Control2 lines (or RF_Out1 and RF_Out2). If I read the datasheets correctly, we have:

• Control1 = 0V, Control2 = -5V: RF_Out1 active on SSW-124, RF_Out2 active on MSWA2-50+
• Control1 = -5V, Control2 = 0V: RF_Out2 active on SSW-124, RF_Out1 active on MSWA2-50+

Since you wired RF_Out 1 on the SO8 to RF_Out1 on the QFN, and similarly for RF_Out2 as well as the Control lines, you need reversing them, don't you?

Again, I might mistaken, but please check. 

[KungFuJosh]

Somebody else please verify.

I might be getting this wrong in my haste, but according to the schematic you posted you may need to swap the Control1 and Control2 lines (or RF_Out1 and RF_Out2). If I read the datasheets correctly, we have:

• Control1 = 0V, Control2 = -5V: RF_Out1 active on SSW-124, RF_Out2 active on MSWA2-50+
• Control1 = -5V, Control2 = 0V: RF_Out2 active on SSW-124, RF_Out1 active on MSWA2-50+

Since you wired RF_Out 1 on the SO8 to RF_Out1 on the QFN, and similarly for RF_Out2 as well as the Control lines, you need reversing them, don't you?

Again, I might mistaken, but please check. 

Good idea to check that. That's much more reading than I did. Seems I need to study those a docs a little more.

[KungFuJosh]

I'm not sure. This needs translation.

It sounds like you're probably correct, but it's slightly ambiguous.

[rf-messkopf]

I'm not sure. This needs translation.

It sounds like you're probably correct, but it's slightly ambiguous.

Datasheet exegesis can be a source of true joy. This is one of the things you'd normally fix in software in your own designs. Anyway, as I read it, the two devices behave in opposite ways with respect to Control1 and Control2, so the lines should be reversed, but that doesn't necessarily mean anything. Maybe we can get another opinion. 

Anyway, it's good to have a way to replace that switch. This won't be the last time one of those turds will fail in an SML, so thank you for designing it.

Perhaps you can panelize that small board somehow as a breakout board. That would make handling and soldering much easier. A FR4 thickness of 0.5mm would probably be ideal.

[TylerPeppy]

I'm not sure. This needs translation.

It sounds like you're probably correct, but it's slightly ambiguous.

Just having a quick look at your schematic versus the datasheets:

For MSWA2-50+:
1.) When VC1 = 0V, VC2 = -5/-8V => RFO1 = OFF, RFO2 = ON
2.) When VC1 = -5/-8V, VC2 = 0V => RFO1 = ON, RFO2 = OFF

For SSW-124:
1.) When VC1 = 0V, VC = -5V => J1-J2 path ON, J1-J3 OFF
2.) When VC2 = -5V, VC2 = 0V => J1-J2 path OFF, J1-J3 path ON

On your board:

RFI1 is connected to SOP8 pin 2. RFO1 is connected to SOP8 pin 5, RFO2 is on SOP8 pin 8.

- Looking at MSWA2-50+, the first case turns on RFO2, which means that SOP8 pin 2 is connected to SOP8 pin 8. In the second case, RFO1 is on, which connects SOP8 pin 2 to SOP8 pin 5.

- Looking at the SSW-128 Datasheet, the first case turns on J1-J2 path, connecting SOP8 pin 2 to pin 5. In the second scenario, J1-J3 path is connected, and thus SOP8 pin 2 is connected to SOP8 pin 8.

Assuming your PCB footprint for SOP8 is the same pin numbering as the SSW-124 part, I *think* the control lines probably need to be reversed.

Here's my interpretation:
╔════════╦════════╦════════════════╦════════════════╗
║                ║                ║ SSW-124                  ║ MSWA2-50+              ║
╠════════╬════════╬════════════════╬════════════════╣
║ VC1         ║ VC2          ║ Connected Path         ║ Connected Path         ║
╠════════╬════════╬════════════════╬════════════════╣
║ OFF         ║ -5/-8V      ║ SOP8 (P2 - P5)          ║ SOP8 (P2 - P8)          ║
╠════════╬════════╬════════════════╬════════════════╣
║ -5/-8V     ║ OFF          ║ SOP8 (P2 - P8)          ║ SOP8 (P2 - P5)          ║
╚════════╩════════╩════════════════╩════════════════╝



- Tyler

[KungFuJosh]

Ok, here it is with the control pins reversed. Board thickness reduced to 0.8mm.

I don't love that I needed to use vias with uneven tracks, but it's a switch, so I don't think it should matter so much. Anyway...

USE AT YOUR OWN RISK.

Thanks,
Josh

[TylerPeppy]

Ok, here it is with the control pins reversed. Board thickness reduced to 0.8mm.

I don't love that I needed to use vias with uneven tracks, but it's a switch, so I don't think it should matter so much. Anyway...

USE AT YOUR OWN RISK.

Thanks,
Josh

Looks fantastic! I was wondering for mfg, if a flex PCB would be better thickness wise. It would probably be more susceptible to vibrations, but could be an option.

[KungFuJosh]

I wouldn't go nuts with the PCB type. Flex PCB adds extra risks like copper tear out.

0.8mm FR4 should be fine. I would use ENIG and maybe a heavier weight, with a minimum of 25um for hole copper thickness.

[squadchannel]

JLCPCB can't (too small for castellated holes - they require minimum 10mmx10mm).

At JLC, you don't necessarily need to enable the "Castellated Hole" option.
https://www.eevblog.com/forum/manufacture/manufacturing-less-than-half-castelated-holes-in-jlcpcb/msg6221843/#msg6221843
I've made a few PCBs in the past that violated the design rules for castellated holes, but there weren't any major issues, and JLC didn't say anything about it.

[KungFuJosh]

At JLC, you don't necessarily need to enable the "Castellated Hole" option.

I've made a few PCBs in the past that violated the design rules for castellated holes, but there weren't any major issues, and JLC didn't say anything about it.

That's good to know, but I avoid gambling with JLC these days. PCBWay seems to have better pricing in this case anyway.

[KungFuJosh]

PCBWay said:

Your design includes half-holes, but the portion of the half-hole inside the board edge is less than 0.25mm. To meet our manufacturing requirements, we recommend increasing the hole diameter or adjusting the hole position to ensure there is at least 0.25mm of the hole remaining inside the board. 

I will fix that and upload a new gerber set.

Thanks,
Josh

[KungFuJosh]

Here's the updated gerber set.

USE AT YOUR OWN RISK.

Thanks,
Josh

ETA: This version passed review at PCBWay.

[tszaboo]

I got myself an SML02 with some issues. So far:
1. Radom screw loose in the enclosure on arrival, falls out at the back.
2. Output level not even close, 30dB or more difference from set level and output. At least the output goes to 2.2GHz. So I'm guessing blown output amplifier. Probably also issues with some of the switch MMICs.
3. Lines on the LCD. Probably a zebra strip issue, I have to se if this can be repaired or not.
4. This on is strange, very slow response to button presses. Often times I have to press a button for 2 seconds for it to register. The encoder is fast though, so not sure what's going on.

I knew about 2 and 3 when buying, let's see if I can repair this.

[KungFuJosh]

A good cleaning and a firmware update might help with the responsiveness. From the sounds of it though, you probably need to do a thorough cleaning and inspection of the mainboard also...and check the PSU voltages.

What options are installed?

[TylerPeppy]

I got myself an SML02 with some issues. So far:
1. Radom screw loose in the enclosure on arrival, falls out at the back.
2. Output level not even close, 30dB or more difference from set level and output. At least the output goes to 2.2GHz. So I'm guessing blown output amplifier. Probably also issues with some of the switch MMICs.
3. Lines on the LCD. Probably a zebra strip issue, I have to se if this can be repaired or not.
4. This on is strange, very slow response to button presses. Often times I have to press a button for 2 seconds for it to register. The encoder is fast though, so not sure what's going on.

I knew about 2 and 3 when buying, let's see if I can repair this.

You might want to read about the tests I did on mine and how the output unlevel issue was interestingly just a module. You didn't mention if it has any other installed options though, so my experiences might not be worth much.

On the other hand, can you set up a slow sweep from 9kHz - 2.2GHz and leave trace persistence ON on the SA for the test? I am curious to see if the gain error is flat.

The other thing, is you might want to do a self-test. If it passes, the output amp is probably fine, and something else in the path is broken. Mine did this, and all self-tests passed but the B3 pulse mod option was busted. When bypassed, it had no issue on the output.

[KungFuJosh]

Probably worth trying self-calibration also.

[TylerPeppy]

Probably worth trying self-calibration also.

Self cal might find some issues, but probably won't solve a 30dB discrepancy.

Also, read the repair manual, and in specific, section 7.6. It has the test point voltages for the 2/3GHz output module.
(https://manuals.repeater-builder.com/te-files/R&S/R&S%20SML_SHB_BD2_03%20Service%20part%202.pdf)

You should be able to access the test point utility through the front panel interface to check the voltages as well. This was useful when checking my output from the B3 module.

Tyler

[tszaboo]

I got myself an SML02 with some issues. So far:
1. Radom screw loose in the enclosure on arrival, falls out at the back.
2. Output level not even close, 30dB or more difference from set level and output. At least the output goes to 2.2GHz. So I'm guessing blown output amplifier. Probably also issues with some of the switch MMICs.
3. Lines on the LCD. Probably a zebra strip issue, I have to se if this can be repaired or not.
4. This on is strange, very slow response to button presses. Often times I have to press a button for 2 seconds for it to register. The encoder is fast though, so not sure what's going on.

I knew about 2 and 3 when buying, let's see if I can repair this.

You might want to read about the tests I did on mine and how the output unlevel issue was interestingly just a module. You didn't mention if it has any other installed options though, so my experiences might not be worth much.

On the other hand, can you set up a slow sweep from 9kHz - 2.2GHz and leave trace persistence ON on the SA for the test? I am curious to see if the gain error is flat.

The other thing, is you might want to do a self-test. If it passes, the output amp is probably fine, and something else in the path is broken. Mine did this, and all self-tests passed but the B3 pulse mod option was busted. When bypassed, it had no issue on the output.

I did read through this thread. I didn't see any options listed at the back, not sure if RS does that. And I didn't have time to take it apart. That have to wait till this weekend. I hope there is an option, seemingly it uses the same switch MMICs so I would rather have a working output level than pulse generation.
I also made a quick scan, measuring the output level at different frequencies and it seemed mostly consistent. I've also saw sometimes 2nd order harmonics (as  I recall) 20-30 down, which would mean I think that filter is getting bypassed, maybe a stuck MMIC.
Shahriar from TSP had a few repairs where it was the output cabling, I don't think I'm that lucky, but we will see. The test you suggests make sense.

Also, when looking at this thread before buying the SML02, I noticed that the manual's block diagram and the teardown photo KungFuJosh posted, they don't seem to align 100%. I'm not sure how exactly, but the switches seem to do something different than what's in the manual.

[rf-messkopf]

4. This on is strange, very slow response to button presses. Often times I have to press a button for 2 seconds for it to register. The encoder is fast though, so not sure what's going on.

This is a known problem with older SML. From what I hear it is the contact pill in the elastomeric keypad that degrades. I know that some people have experimented with conductive silver, but that is apparently difficult to fix. Luckily, mine doesn't do that.

[TylerPeppy]

I got myself an SML02 with some issues. So far:
1. Radom screw loose in the enclosure on arrival, falls out at the back.
2. Output level not even close, 30dB or more difference from set level and output. At least the output goes to 2.2GHz. So I'm guessing blown output amplifier. Probably also issues with some of the switch MMICs.
3. Lines on the LCD. Probably a zebra strip issue, I have to se if this can be repaired or not.
4. This on is strange, very slow response to button presses. Often times I have to press a button for 2 seconds for it to register. The encoder is fast though, so not sure what's going on.

I knew about 2 and 3 when buying, let's see if I can repair this.

You might want to read about the tests I did on mine and how the output unlevel issue was interestingly just a module. You didn't mention if it has any other installed options though, so my experiences might not be worth much.

On the other hand, can you set up a slow sweep from 9kHz - 2.2GHz and leave trace persistence ON on the SA for the test? I am curious to see if the gain error is flat.

The other thing, is you might want to do a self-test. If it passes, the output amp is probably fine, and something else in the path is broken. Mine did this, and all self-tests passed but the B3 pulse mod option was busted. When bypassed, it had no issue on the output.

I did read through this thread. I didn't see any options listed at the back, not sure if RS does that. And I didn't have time to take it apart. That have to wait till this weekend. I hope there is an option, seemingly it uses the same switch MMICs so I would rather have a working output level than pulse generation.
I also made a quick scan, measuring the output level at different frequencies and it seemed mostly consistent. I've also saw sometimes 2nd order harmonics (as  I recall) 20-30 down, which would mean I think that filter is getting bypassed, maybe a stuck MMIC.
Shahriar from TSP had a few repairs where it was the output cabling, I don't think I'm that lucky, but we will see. The test you suggests make sense.

Also, when looking at this thread before buying the SML02, I noticed that the manual's block diagram and the teardown photo KungFuJosh posted, they don't seem to align 100%. I'm not sure how exactly, but the switches seem to do something different than what's in the manual.

This is interesting, when you mention the harmonics. They should not appear that high relative to the carrier. This is almost the exact same issue from my B3 pulse modulation option, where the switch MMIC died and one of the output amplifiers was also broken according to test points.

Not entirely sure what components were all broken on the B3, as I bypassed it and called it a day. I am hoping to swap the switches out at some point, though.

[nctnico]

4. This on is strange, very slow response to button presses. Often times I have to press a button for 2 seconds for it to register. The encoder is fast though, so not sure what's going on.

This is a known problem with older SML. From what I hear it is the contact pill in the elastomeric keypad that degrades. I know that some people have experimented with conductive silver, but that is apparently difficult to fix. Luckily, mine doesn't do that.

You can buy 'stickers' from Buttonworx which have a conductive pattern printed on them. They go between the original rubber key mat and the contact area. They work like a charm.