JYE Tech DSO150 oscilloscope troubles

[MrElectric]

Just for fun, I bought a JYE Tech DSO150 oscilloscope from Banggood. I thought it would be a fun soldering project that might be usable sometimes.

When it arrived, I read the instruction, which says to verify that it boots before soldering anything. Else the warranty will be void. I found a 9V power supply with center positive as instructed.

I connected the power. The screen gets light, but nothing happens. After a while, I smell that something is getting hot. I remove the power and can feel that the large IC is really hot. I first thought the power supply might be faulty. So, I connect my power supply instead. Exactly 9V. I didn't think to turn down the current limit on the supply. After turning it on, it lights up exactly like when I used the power supply. Light screen, and a red diode on the backside. Then I notice that it delivers 1.5A through the circuit, and I see some smoke after a few seconds. According to the manual, it needs 400mA.

I reported this to both Jyetech and Banggood. Jyetech send me a replacement. Banggood sent me another complete kit because I had received a bad unit.

The new ones arrived yesterday.

I first tried the one Jyetech sent me (with the same power supply as last time - I have verified it since). Lo and behold, it boots up, and the screen looks as expected. I try the same thing with the other one, but it's as dead as the first one I received.

I now try the Jyetech one once more, and now it won't boot again...

So, I set up my power supply again. This time, I use the current limiting. I set it at 200mA first. I test all three boards, and none work. I up the current to 400mA. Now the red diode glows dimly on one, but not on the others. I up the current to 600mA, and now the red diode shines brightly.

As I didn't have a schematic, I did do some measuring.

The input is 9V, and then there's a voltage regulator which outputs 3.3V. There is a pad that says 3.3V. This looks like a good spot to measure the voltage. I can see no consistent results between the three boards. None of them measure 3.3V there though.

Using the continuity tester, I see that there is a short between all legs of the voltage regulator and ground on two of the boards. The third has 174 ohms between ground and 3.3V, while the others have 0.5 ohms (or thereabouts). Clearly, there is a short somewhere, that makes the voltage regulator get hot.

On one of the boards, it's the big IC that gets hot, while on the other two, the voltage regulator gets hot...

Anyone else with this problem? I seem to find a number of positive reviews of this unit.

[edit: I removed the negative comments. I understand more of the circuit now, and believe the margin for over-voltage wasn't enough. Many power adapters give out much more than 9V when unloaded. The voltage regulator only handles up to 15V. They are going to release a new version with a 5V USB power connector instead, which will make for a better tolerance to over-voltage]

[Peabody]

I built the Shell scope, and it has worked very well.  From what I hear, I think your experience is not at all typical.  In fact, having the same failure on three units from different sources would be extremely unlikely.  From your description it seems the only thing in common in all three cases is your power supply.  Have you considered trying it with a 9V battery?

My DSO150 is the original version, and it draws about 120ma.  They've made some improvements since then, and I don't know what the newer units draw.

I really don't think these units are junk at all.  A good bit of engineering goes into them, and the manuals are well done.  And they have continued to issue firmware updates from time to time, which is unusual for products of this type, and as you have seen, they stand behind their products.  JYETech is about to release a two-channel version of the DSO150, and if the price isn't too high, I plan on getting one.

Of course I can't tell from here, but logically it seems something at your end is more likely the problem.

By the way, there should be a schematic included with the unit.  If not, schematics are available for download on the JYETech.com website.

[MrElectric]

I found this instcruction, where someone has similar problems (after using a 12V AC adapter):
https://www.instructables.com/id/DIY-Oscilloscope-Kit-Assembling-and-Troubleshootin/

I now see I wrote "power supply" everywhere yesterday. I need to clarify that. When I got the first one, I first couldn't find a 9V power adapter, so I used my lab power supply, set to exactly 9V (tested with a multimeter as well). It had the same symptom then, i.e. white screen, red diode. I only did this for a short while, because I didn't find a place where I could attach the +9V clip, so I had to hold it in place while testing. I then found a 9V power adapter that I used after that. Same symptom. After a while the large IC got hot. I then switched back to the lab power supply because I wanted to be sure I supplied the correct voltage. I left it running like this for 10-15 seconds, when some component (maybe the voltage regulator) blew up.

When I received the new units, I used the 9V power adapter. The first one booted up ok, and I don't think that anything got hot. The second one didn't boot up at all, same problem as unit 1. Next time I tried unit 2, it didn't boot up.

I have since then tested with the lab power supply.

I have now found the schematic online. It was not included in any of the units.

[Peabody]

Well it just seems unlikely that three units would be bad.  As you said, the reviews are generally quite good.  But as confirmed on the link you found, powering it with higher than recommended voltage would be a problem.

I'd strongly suggest you post about this on the JYETech forum, and see if anyone there has a suggestion.  It sounds like a voltage regulator problem, but of course it could be something shorting downstream.  Nothing on the board should get hot, and it shouldn't draw much over 100ma.  I've converted mine to battery power with a charger and boost converter, and it still works fine after two years.  I have found it to be surprisingly useful.

My unit included a schematic, but they may have stopped doing that.  The printed version was difficult to read, and I guess they decided the downloaded pdf would be more useful.

I should say that there are some counterfeit units out there, but Banggood is known to have genuine supply.  I got mine there.

I hope you find a resolution.

[NoisyBoy]

I received mine last week and just assembled it last night, everything works fine on it, no problem.

I also think with three failure, it is likely that the problem could be related to power supply.

I have powered mine with both an Agilent bench supply and a precision system power supply, the latest draws just about 100mA in normal operation.  With such project, I always use a 300mA current limit to be on the safe side.  I am surprised that the tiny scope did not burst into flame when it drew 1.5 A

[mikerj]

What kind of "9v" power adapter did you use?  If it was a basic unregulated transformer type then the off-load voltage would likely have been way above the maximum for the DSO150.

[rsjsouza]

Looking at the schematics, I can't help but wonder if the voltage regulator was at fault. The main IC is most probably operating a 3.3V and the LM1117/3.3 voltage regulator should be fine for up to 15V. 

Also, IIRC several 9V plugpacks from the 1980s put out AC instead of DC. I am not sure if this is still a thing these days, but it certainly was a source of headaches. Given the DSO150 has a 1N4007 and two 10uF capacitors, it will certainly put DC but with a massive ripple.

i am basing my assumptions on the schematics below:
https://jyetech.com/wp-content/uploads/2018/07/dso150-shell-schematic-main.pdf

[MrElectric]

On one of them, I de-soldered all capacitors and ICs, and finally found that it was the large IC (STM32 F103C8T6) that was shorting 3.3V to ground.

I removed the voltage regulators from both this, and the board without the short.

I did some testing, and found that the voltage regulators are outputting the input voltage, instead of 3.3V. That means they are both fried (probably the third one as well).

So, I guess that an over-voltage fried the voltage regulator, and, when it stopped regulating and delivered 9V+ to the STM32, it fried the STM32 as well.

One could hope that there was some protecting circuit, to protect the STM32 from over-voltage. For example a voltage regulator that doesn't short when fried.

[rsjsouza]

Thanks for the details; that confirms my suspicion and, with the massive number of falsified devices on the Chinese market, I am not surprised at all that JyeTech may have been caught by surprise with a batch of crap LM1117s.

Perhaps I would contact the seller and JyeTech with this information; it may save a lot of grief for other developers out there eager to have a DSO on their hands.

[Gyro]

Failed dodgy regulators (78L05) were an issue on the DSO138 too, as were fake TL084s irrc. They've been badly hit by clones of their various models too.

It must be hell being a small Chinese manufacturer having to buy jellybean parts on the local market and then getting ripped off by your fellow countrymen.

[rsjsouza]

Failed dodgy regulators (78L05) were an issue on the DSO138 too, as were fake TL084s irrc. They've been badly hit by clones of their various models too.

That is the cloning game: "just give me the sample", as it was common a decade or two ago (it probably still is).

Sometimes it is very difficult to distinguish a clone: I got a LA2016 (still on the boat) bought from an Aliexpress seller named kingst Official Store, but I then noticed the screenshots do not mention Kingst but Jiankun. They may be from the same people, but how the heck do I know?!?

It must be hell being a small Chinese manufacturer having to buy jellybean parts on the local market and then getting ripped off by your fellow countrymen.

I have seen a lot of these happening in Brasil... In late 1980s there was a brand named "Brokosonic" that ripped some designs from other companies (mostly AM/FM and cassete players). In the early 1990s, "Bakosonic" came to be, ripping off "Brokosonic" designs.

[MrElectric]

I'm going to order some replacement components. I'll order the STM32F103CBT6 instead of the STM32F103C8T6 (a B instead of an . The STM32F103C8T6 seems to be hard to get today, and from what I can see, the only difference is that the STM32F103CBT6 has more memory (128kB instead of 64kB).

I also found the 1117 in different versions. There's a cheap one, made by Diodes incorporated in Malaysia (ZLDO1117K33TC) and a slightly more expensive one by Texas Instruments (LM1117MPX-3.3/NOPB). I'll order the Texas Instrument one. The pinout is the same for all of these, but I trust TI more than an unknown Malaysian company (though it's probably much better than the chinese one on the PCB (AMS1117 3.3 HK843F)),

I'll order an Uart-USB converter as well, so that I can program the new CPU's.

[mikerj]

FWIW Diodes Inc is a large and well respected American semiconductor company, they make a huge range of parts.  No need to avoid their parts (in fact I've had more issues with TI parts than Diodes Inc).

[Peabody]

MrElectric, I'm not sure it makes sense cost-wise to order all those parts.  Even if you replace them, it still may not work.  I saw that you have a conversation going on with JYETech on their forum.  I would bet that if you contact them privately, they would send you a new unit at no cost except whatever duty you have to pay.

[rsjsouza]

Peabody, I would agree with you if it weren't for the first sentence on MrElectric's post:

Just for fun, I bought a JYE Tech DSO150 oscilloscope from Banggood. I thought it would be a fun soldering project that might be usable sometimes.

I think it is the journey that counts in this case. 

[HwAoRrDk]

A lot of these cheap 1117 voltage regulators only have an absolute maximum input voltage rating of 15V. An unregulated 9V plug pack could easily exceed that under light load. For example, a random multi-output one grabbed from my shelf gives, with no load, nearly 16V on the '9V' setting and just over 20V on the '12V' setting. I have a DSO138 and the unregulated plug pack I run it off I have set to 6.5V, which actually delivers it about 8V.

It is my experience that these cheap 1117s fail with the input shorted to output when over-volted. And STM32s definitely do not appreciate 15V up the clacker.

[NoisyBoy]

The sensitivity to over voltage is a well known issue, that’s why they have caution about keeping it at 8-10V all over the latest documentation. 

Even with a bench power supply, I would be careful with the cheap supplies due to overshoot at power on.  I know with a very low ripple linear power supply, the DSO150 actually behave quite nice. 

[MrElectric]

Those parts are quite cheap. I think I paid less for them than what I have to pay for the customs and postal handling fees for a china parcel.

I actually paid a lot for these four(!) parcels already. I'd rather pay for new components than more senseless fees.

Also, I did email about the first unit, and was sent a new one from Jye Tech (without charge). But it still cost me to receive the unit.

And it's more fun to try to fix the ones I have!   

[MrElectric]

FWIW Diodes Inc is a large and well respected American semiconductor company, they make a huge range of parts.  No need to avoid their parts (in fact I've had more issues with TI parts than Diodes Inc).

Ok. I had a look at their datasheet, and saw that their 1117 can handle 1A and up to 18V, while other manufacturers' 1117's only handle 800mA and 15V. And they only cost about a third of the TI parts. Unfortunately, I have already ordered, so I will have to do with the TI ones. :-)