The Sinclair QL Forum

Today I decided to built the hardware needed to equip my Q68 with QL network ports. I had procured the parts many weeks ago from a local shop. However although I ask specifically for a BAT43 diode, I just realize the clerk gave me 1N5819 diodes instead. Equivalent parts ? When I look at the specs, they seem quite different!

Can someone more knowledgeable confirm if this would be acceptable ?

Thank you!

François

I‘m not an electronics guy, but looking at the specs I think you can safely substitute the 1N for the BAT (but not the other way round) because it has higher specs. Only downside is that it‘s a bit bigger.

Hi,

I have built many Q68 QLNET adapters, using the BAT43 diodes, which the QLNET connection worked first time.

I can only suggest either replace the 1N5819 with the correct BAT43 diode or try the 1N5819 diode, it may work, but no guarantees.

Derek_Stewart wrote:Hi,

I have built many Q68 QLNET adapters, using the BAT43 diodes, which the QLNET connection worked first time.

I can only suggest either replace the 1N5819 with the correct BAT43 diode or try the 1N5819 diode, it may work, but no guarantees.

Not working is one thing, but damaging the Q68 is another. If you think there is no risk then I will try and report.

François

Despite my caveat I‘m 99% sure it will work as a replacement, even without having seen the schematic, and 99.9% sure it won‘t damage anything

mk79 wrote:Despite my caveat I‘m 99% sure it will work as a replacement, even without having seen the schematic, and 99.9% sure it won‘t damage anything

99.9% sure! Ok good enough for me

Checked the schematics now. If I read that correctly it‘s a poor man‘s level shifter using the diode as a voltage clamp. As the 1N has a lower forward voltage I think it‘s an even better protection for the Q68 pin than the BAT.

FrancoisLanciault wrote:

mk79 wrote:Despite my caveat I‘m 99% sure it will work as a replacement, even without having seen the schematic, and 99.9% sure it won‘t damage anything

99.9% sure! Ok good enough for me

Just be careful, it is all well quoting data sheets, but implementation is another thing.

There should be care taken connecting the Q68 expansion port, which I did not fit a 2x12 header on the Q68 till late in Batch 2. This is because I did not think there would any extra hardware required.

How wrong, when Martyn wrote the QLNET driver, I started to fit a header on the Q68 expansion port.

FrancoisLanciault wrote:However although I ask specifically for a BAT43 diode, I just realize the clerk gave me 1N5819 diodes instead. Equivalent parts ?

My problem with the 1N5819 is that I find no specification for the speed yet. The BAT43 is a relatively fast (5 ns) small signal schottky diode, whereas the 1N5819 is a power diode typically for rectifying purposes with larger currents. The diode should be <10 ns fast to sufficiently protect the input of the FPGA.

Peter wrote:

FrancoisLanciault wrote:However although I ask specifically for a BAT43 diode, I just realize the clerk gave me 1N5819 diodes instead. Equivalent parts ?

My problem with the 1N5819 is that I find no specification for the speed yet. The BAT43 is a relatively fast (5 ns) small signal schottky diode, whereas the 1N5819 is a power diode typically for rectifying purposes with larger currents. The diode should be <10 ns fast to sufficiently protect the input of the FPGA.

Here I see 10ns

https://www.digikey.com/product-detail/ ... -ND/950588

A bit too close to the acceptable speed ?

Francois