# Problem between DSP and PYNQ-Z2

**URL:** <https://discuss.pynq.io/t/problem-between-dsp-and-pynq-z2/4800>\
**Category:** Support\
**Created:** [October 18, 2022, 2:27am UTC](https://discuss.pynq.io/t/problem-between-dsp-and-pynq-z2/4800 "2022-10-18T02:27:11Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![hsuharry12](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/hsuharry12/32/3715_2.png) [@hsuharry12](https://discuss.pynq.io/u/hsuharry12)\
**Post date:** [October 18, 2022, 2:27am UTC](https://discuss.pynq.io/t/problem-between-dsp-and-pynq-z2/4800/1 "2022-10-18T02:27:11Z")

</div>

Continuing the discussion from [Spi on PYNQ-Z2 Pmod](https://discuss.pynq.io/t/spi-on-pynq-z2-pmod/4633/10):

The pynq-z2 board can use slave now, but there is a problem when communicating with DSP. We initially guessed that there would be one more 0 at the front. For example, if I originally wanted to transfer 1111111111111111, the result would be 01111111111111111. The redundant 1 would be pushed to the next column, but after trying to push all of them forward manually, there still be some errors, so I don’t know what caused the problem.

---

<div class="post-metadata">

**Author:** ![mizan](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/mizan/32/2119_2.png) [@mizan](https://discuss.pynq.io/u/mizan)\
**Post date:** [October 18, 2022, 4:47am UTC](https://discuss.pynq.io/t/problem-between-dsp-and-pynq-z2/4800/2 "2022-10-18T04:47:25Z")

</div>

As i mentioned earlier in previous post, the slave mode has a bug (i think) which provide inconstitant data at beginning which i could not solve either, sometimes it adds 1 cycle, sometimes 2 and sometimes none. I don’t think it is problem with the DSP, more like it is related with the IP itself.

---

<div class="post-metadata">

**Author:** ![hsuharry12](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/hsuharry12/32/3715_2.png) [@hsuharry12](https://discuss.pynq.io/u/hsuharry12)\
**Post date:** [November 2, 2022, 9:41am UTC](https://discuss.pynq.io/t/problem-between-dsp-and-pynq-z2/4800/3 "2022-11-02T09:41:39Z")

</div>

It turns out that our guess is right, but there are some mistakes pushing the data manually in the Python code. It works fine now although the bug still exists.