Problems with TLE5206 H-Bridge

BillBWann

Member
I’m wondering if anyone else has used a TLE5206 H-Bridge recently and what is their experience with them. I first bought 4 of these from China back in 2012 and found them very useful. One was used in a model train with PWM for forward and backward motor control and another to drive an electric drill to raise and lower a chook door. Both are still working flawlessly. I bought another 4 in 2013 again from China to use as spares for later.

Recently, I was building a very simple power supply which was to use an 08M2 and the TLE5206 H-Bridge but the H-Bridge didn’t behave as expected and appeared to cause problems to the USB to Serial RS232 cable device (also from China) when programming the 08m2. The software kept losing the device and the USB had to be taken out and reinserted to get recognised again. Eventually, it failed completely and couldn’t be recognised. The output from the H-Bridge was also erratic and output 1 (pin 1) would go high when both inputs 1 & 2 (pins 3 & 5) were low but would go low with any other combination of inputs. In fact, with both inputs high, both outputs were low although output 1 did rise to a couple of volts. Output 2 (pin 7) appeared to operate normally and follow input 2 (pin5) as long as input 1 (pin 3) was kept low.

Summarising
Code:
    Input       Output
pin3	pin5	pin1	pin7
 L       L       H      L
 L       H       L      H
 H       L       L      L
 H       H    2.14v  0.5v
(apologies about the formatting. I don't know how to improve it)

I assumed that I’d somehow destroyed the TLE5206 H-Bridge by inadvertently touching something that I shouldn’t so I replaced it only to find the circuit still behaved exactly as before. I again suspected that it was something I’d done and made some blunder with my circuit board but careful inspection of it didn’t reveal anything.

I then made up a very simple circuit with an 11 volt battery supply (3 Li-ion batteries) to pins 6 & 4 (also with a 220uF capacitor across them) and with both inputs grounded with a 1k resistor. I placed a LED & 1k resistor across the outputs and as you can see from the picture, the LED lit up when both outputs should be low.

IMG_5891.JPG

I got exactly the same result from every other TLE5206 that I had. I then ordered another 6 TLE5206’s from a different supplier in China which have just arrived and they are all showing this same odd behaviour.

It seems extremely unlikely to me that of the 14 devices that I’ve now purchased from 2 different suppliers (and also the latest 6 are the 2S version of the device while the earlier 8 were the 2G version), the first 3 (picked at random) worked perfectly but all the remaining ones are flawed in the same way.

If anyone can see that I’m doing something stupid or I’m overlooking something obvious, please let me know as I’ve now spent a considerable amount of time checking this out. Also if anyone has purchased any of these recently and they are working as expected, then please let me know your supplier.
 
Last edited by a moderator:

premelec

Senior Member
Perhaps download an actual schematic as copied from your breadboard - and program listing... hard to see what you are doing from your BB picture....
 

hippy

Technical Support
Staff member
It does seem odd, especially that some work but many don't. It is possible that some dodgy chips have entered the market and it could be possible multiple sellers have become a victim of that.
 

binary1248

Senior Member
Your truth table certainly looks odd. If you publish your complete schematic it would help clarify your problem.
Here is the pinouts, same as yours ? And here is a truth table from the data sheet.
.
Also what does the error flag output on the device (pin 2) indicate on each input state ?
.

52065_pinout.jpg52065 Truth_tbl.jpg
.
Trying to look closely at your BB, it looks like pin 1 (out) is grounded ??
 
Last edited:

fernando_g

Senior Member
The output truth table is indeed odd.

The last state in particular, H,H = 2.14, 0.5 is indeed strange.
Do you have a BTL load?
 

BillBWann

Member
Thanks for the replies.

My test circuit doesn’t involve a picaxe and is the simplest one I could think of. I used it because I was concerned that my original circuit was somehow destroying the H-bridge because it seemed very unlikely that the 2 H-Bridges that I used in that circuit could have both been faulty when I had had no previous problems with them.

The circuit was so simple that I thought a description was sufficient but for the record, it was as sketched below.

TLE5206.jpg

Having said that, I now realize that I inserted the wrong photo in my original post and that photo shows another variation that I was also using at one time where output 1 was connected to ground and not Out 2 as I described. Also trying to confirm that I’ve set up the breadboard correctly is made difficult in that photo because the resistors and LED are about an inch above the breadboard and so there are parallax errors but rest assured that I’ve spent a lot of time looking at this over the last 5 weeks or so and I really can’t believe that I’ve never managed to set up one chip correctly.

Output 1 is definitely not connected to ground.

I haven’t been monitoring the error flag continuously but it wasn't pulled low on the few occasions that I did check it. Also I’ve never seen the circuit drawing more current than would be expected from the load imposed across its outputs. Initially, I was using a current limited power supply and I never noticed it going into current limiting mode. I moved over to batteries to rule out the possibility of my power supply causing a problem as the device seems to need a pretty large capacitor on its supply (but presumably that for when using PWM).

The points I was trying to make in my original post were to get confirmation from forum members that:

• Out1 was high when In1 was low which was opposite to that specified in the datasheet and this situation would be impossible to achieve with a proper functioning chip and this circuit.

• That it was unlikely that my test circuit could have been destroying all of my stock of TLE5206’s.

• That there wasn't a simple scenario that could explain why such an improbable situation could have arisen with so many faulty chips after the first 3 had functioned so well.


The chip pins are basically symmetrical and the only real way to connect it up incorrectly is to apply battery voltage to the EF output. As its an open collector output, I would doubt if even that could cause any real problem and I certainly haven’t done that to every chip that I’ve checked.

I have checked what the chip does when a load is applied and it gets even weirder and I won’t bother describing it in detail as it does vary marginally from one chip to another and also on the current being drawn. Suffice to say that output 1 often doesn’t rise to the full 11 volts but seems to settle at about 3 different voltages (2.5, 5 and 11 volts) and it seems to select one of these voltages seemingly at random after an input has changed or sometimes a couple of seconds after an input has changed. The voltage appears to be largely independent of load current once a small load has been imposed.

The output on out 1 is also influenced by input 2 (or maybe output 2) because, with both inputs low, out 1 is high as shown in the photo but it goes low as soon as input 2 is taken high (by high, I mean to 5 volts, not 11 volts) which suggests an internal short of some sort between the 2 arms of the H_Bridge.

When the TLE5206 works, it’s a very useful device as it has high current capabilities (up to 5 amps continuous), short circuit protection and contains the freewheeling diodes for inductive loads – and all in a small convenient single component that costs about $5 from China. I’m somewhat surprised that it’s not more popular on this forum and wondered if it’s because it had a history of not working as specified but I couldn’t find anything on Google that suggested that.

Based on the responses so far, there would appear to be little to no practical experience with this chip by forum members which surprises me for a device that has so much potential. I really would like to find a supplier of functioning chips like the first 3 I used.
 

hippy

Technical Support
Staff member
Are you sure you actually have TLE5206's and not TLE5205's which would seem to have a behaviour closer to what you describe ?

I would guess the lack of apparent use amongst PICAXE members would come from simply not having had a need to use such a component, using lower current L293D or some other alternative, or members who have used them are not visiting the forum right now.

You might be able to get more help from a robotics forum where H-bridge drivers are likely used more often, and Infineon who manufacture the device may be able to help, or members of their forums.
 

binary1248

Senior Member
As I follow this, hippy sure has a good point, here is the truth table for 5205
interresting.jpg
Looks the same as what BillB published.
 

BillBWann

Member
Are you sure you actually have TLE5206's and not TLE5205's which would seem to have a behaviour closer to what you describe ?
Thanks Hippy & Binary for your replies.

Once again Hippy I’m impressed by your ability to think laterally and hone in on a problem. I had no idea such a device existed and it seems to be a very useful device.

IMG_5895.JPG

As you can see from the photo, both types of TLE5206 chips that I have are clearly marked 5206 but the latest 2S version certainly behave like a 5205 and may even be a better choice for what I was planning to use the 5206’s for - with their high impedance ability when in the PWM off state. Getting PWM with high impedance during the off period in both directions may be a bit tricky though and I’ll have to give that a bit more thought.

This possible mix-up in chip markings doesn’t explain the peculiar performance of my earlier 2G versions which still seem to have stability problems but they also seem to operate more like 5205’s than 5206’s.

Thanks again. It now looks like I have at least 6 chips that will probably be able to achieve what I want.
 
Top