[lucas]

I can drop down that powersupply to min 41V and I will try this.
Motor current had been selected to 5 A and I tried some quick stepper movings with quick
acceleration too.
But should it be better to isolate TQ from the circuit when connecting to LATCH or not?

These drives: TB1H and HY-DIV268N are offered on Ebay spec'd at 5A current and 50V supply, this is simply cheating and lying.

5A is ALSO the absolute maximum peak value for 100msec.
Max. current in operating range is 4.5A so you were really over the operating values with 48V and 5Amp.

And these drives won't even handle the operating values if the chip is badly mounted (see pics in the early posts) where burrs on the mounting holes prevent the chip from laying flat on the heatsink resulting in insuficient cooling.

Then there's another "small" problem with these drives where the design doesn't comply with the current datasheet:
http://www.mouser.com/catalog/specsheets/TB6600HG.pdf
On page 8 :
" Pleas use the IC under the conditions as follows; 0.11Ω ≤ RNF ≤ 0.5Ω, 0.3V ≤ Vref ≤ 1.95V "
(This sentence was not present in the preliminary datasheet and other values are changed also.... and I really wonder why Toshiba still doesn't have an official datasheet on their website, weird.....)

Vref should thus be equal or lower than 1.95V, the sense resistor on these drives is 0.22 ohm and at 4.5A current this then results in a Vref voltage of 2.97V wich is far too high.
A drive with 0.22 ohm current sense resistors should thus be limited to 2.95 Amp: 2.95 * 0.22 * 3 = 1.947 volt on Vref.

Reduce the current to 3A max or change the sense resistors and check the mounting holes, remove burrs, clean heatsink and chip, then apply some heatsink paste and assemble as it should.
Bear in mind to handle the drive very gently because the PCB is floating and forces applied to it might damage the pins of the chip.

It's always better to isolate the TQ input but it might be risky to cut traces or remove SMD components without the proper tools.
If your drives match the schematic then there's no need to isolate the pin.

I have a few more issues with this chip, it's datasheet and the available drives but that's for later, don't have the time right now.

Which pins are 3 & 4?

On which side of the chip are they?

Please refer to the datasheet mentioned above, there's a drawing on one of the last pages with the pin numbers.